ProgrammableMarketing/pax.markets (opens in a new tab)
PDF

Views

  • Plan
  • Roadmap
  • Research

On this page

  • The brief
  • Start here
  • Idea portfolio
  • Effort vs impact
  • This week · 8
  • This quarter · 11
  • Big bets · 2
  • Wildcards · 2
  • How we can help
Generated Nov 10, 2026
  1. Programmable
  2. Playbooks
  3. PAX Markets
PAX Markets logo

growth playbook · Nov 10, 2026

pax.markets (opens in a new tab)

PAX Markets growth playbook

Diagnosis

PAX Markets is a six-person YC W25 exchange with unusual technology: the matching engine and traders' logic share one chip. Free-tier traders pay nothing and earn a small rebate, and the company earns money only when market makers pay for λ API speed. The site loads fast and the docs are thorough, but the path to each customer group has gaps.

The speed claim differs between channels (about 40 ns and 500x faster on the homepage, 30 ns and 1000x in the launch post) and nowhere says how it is measured. "Preview markets" opens an alpha app that, without a live data feed, shows an empty order book and a chart error. The homepage's fee example compares against a 10 bps fee, while a new trader at Kraken Pro pays up to 80 bps.

Visitors from New York, Canada or Singapore learn they cannot join only after they try. The Binance-compatible API that would ease switching is "under construction". This playbook fixes those first. Next it builds proof of liquidity (founding market makers, a weekly market-quality report, a λ simulator, client libraries), and later an independent latency audit, an equities ATS waitlist and a licensing option for the silicon.

Growth Score

6/10

Biggest Opportunity

Turn the speed claim into evidence a quant team can check, with one measured number, a method and a sandbox test. That shortens the only sale PAX makes money on, and every maker it wins funds rebates for the free tier.

Biggest Risk

A two-sided market at alpha stage: market makers will not pay for λ without taker flow, and traders will not stay on an empty book. Meanwhile Kraken, Hyperliquid and EDX already have liquidity and trust.

01

Start here

The three moves to make first. Each one is written up in full below.

  1. 1

    Publish one latency number with its method, so "40 ns, 500x" on the homepage and "30 ns, 1000x" in the launch post stop disagreeing

    Decide which number is current, 40 ns or 30 ns, and write down exactly what it measures.

    High impact First result in 2 weeks
  2. 2

    Turn "Preview markets" into a public markets page that shows real specs even when the live feed does not load

    Open "Preview markets" in a private window on a phone and record what shows in the first three seconds.

    High impact First result in 2 weeks
  3. 3

    Show the fee gap in dollars, against what new traders really pay elsewhere

    Collect the current first-tier fees of three rival venues, with links and dates.

    High impact First result in 3 weeks

02

Your idea portfolio

23 ideas, grouped by when to run them. Open any idea for how to run it: the steps, who to approach, copy you can use, and how to measure it.

  • 7 ideas are marked Unexpected: plays you are unlikely to have tried.

    1 of these are things Programmable can build and run for you.

Every idea placed by the effort it takes against the impact it could have. Quick wins sit top left; select a number to open that idea. The filters below narrow the list by status, timing, category and what Programmable can build.

Effort vs impactEach number is an idea. Select one to open it.
Impact →
Quick wins · 4Big bets · 7Fill-ins · 7Reconsider · 5
Effort →

Showing 23 of 23 ideas

This week 8

Quick wins with a clear first action.

  • 01

    PAX sells one thing for money: speed through the λ API, bought mainly by market makers. Those buyers measure latency for a living, and right now they find three different stories. The homepage says the on-chip round trip "takes about 40 ns, 500x faster than the 20 µs" of the fastest traditional venues. The Launch YC post, linked from PAX's YC profile, says "30 nanoseconds (1000x faster)" than a "30 microsecond" benchmark, and a product directory repeats the 1000x. The λ API blog post claims "the fastest tick-to-trade at any venue in any asset class" with no number. No page says what is timed, at which percentile, or on which build (the hiring posts say AWS F2 FPGA today, an ASIC later). To a quant team, a number that changes between pages and has no method reads as marketing, not engineering.

    High impactLow effort First result in 2 weeks Co Founder & CEO

    First action

    Decide which number is current, 40 ns or 30 ns, and write down exactly what it measures.

    How to measure it

    λ sandbox trials sourced from /latency

    Now
    –
    Target
    10 market-making firms
    By
    90 days

    Steps

    1. 1

      Pick the definition

      Name the start and end events of the round trip in plain words, with a diagram.

      Co-Founder & CTO 1 day

    2. 2

      Measure the current build

      Median, p99, p99.9 and jitter on the AWS F2 FPGA build, with the date and version.

      Founding Principal Hardware Engineer 3 days

    3. 3

      Source the comparisons

      Give each external figure (20 µs at traditional venues, 1-10 ms at crypto venues) a public, dated source.

      Co-Founder & CEO 1 day

    4. 4

      Publish /latency

      The number, the method, a histogram, the sources and a changelog.

      Founding Software Engineer 2 days

    5. 5

      Align every claim

      Homepage, Launch YC post, YC profile, λ blog post and directory listings all use the same number and link to /latency.

      Co-Founder & CEO 2 days

    6. 6

      Add "see it yourself"

      Sandbox steps a quant can follow to observe the round trip.

      Founding Software Engineer 3 days

    7. 7

      Use it in sales

      Put the link in the first line of every market-maker email and event follow-up.

      Co-Founder & CEO ongoing

    First result
    2 weeks
    Cost
    $0, staff time only
    Owner
    Co Founder & CEO

    Smallest test

    Today, make the homepage and the Launch post state the same number, and link a one-paragraph method note.

    Why it works

    Publish a /latency page on pax.markets with: one plain definition of the round trip (for example, market event reaching the matching engine to λ order resting in the book); the measured median, p99 and p99.9 on the current FPGA build, with the date and version; how it was measured; a dated, sourced comparison with external order entry at traditional and crypto venues; and a changelog for when the ASIC arrives. Then make the homepage, the Launch post, the YC profile and the λ blog post use the same number and link to the page. Finish with a "see it yourself" section that walks a quant through the self-serve sandbox.

    Market makers choose venues on measured latency and on determinism. A number with a method and a way to test it cuts the diligence time on the one sale that earns PAX revenue. One canonical page also stops directories and AI summaries from copying the old 1000x figure.

    Ready to use

    ChecklistWhat the /latency page must contain

    [ ] One-sentence definition of the round trip (start event, end event)
    [ ] Median, p99, p99.9 and jitter, with build, version and date
    [ ] Measurement setup (clock source, where timestamps are taken)
    [ ] Comparison table: λ on-chip vs external order entry, each source dated
    [ ] Same number on homepage, Launch YC post, YC profile, λ blog post
    [ ] "See it yourself" sandbox walkthrough
    [ ] Changelog (FPGA today, ASIC later)

    Risks and how to handle them

    • A new build measures slower

      Version every number and keep the history on the page; a transparent changelog builds more trust than a fixed claim.

    • Rivals pick the method apart

      Invite review and follow with the independent audit (idea 20).

    Evidence

    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
    • Launch postLaunch YC post (listed as over a year old): "The fastest exchanges today boast a 30 microsecond round-trip latency. PAX crunches that to 30 nanoseconds (1000x faster)"; it claims PAX is faster than NYSE, NASDAQ, Deutsche Börse, EDX, Crossover, Binance and OKX, and describes the model as selling speed and buying order flow.Source
    • Directory listingsYC company profile: "Co-located trading facility and matching engine on one chip", describing deterministic nanosecond tick-to-trade latency; W25, founded 2024, team of 6; it links a LinkedIn company page and no X account. A third-party product directory repeats that PAX "reduces exchange response latency by 1000x".Source
    • Your websiteBlog: four posts, from "In Crypto, Fees are High and Speeds are Slow" (November 6, 2024) to the PFOF post (February 2025), none since. The λ API post claims "the fastest tick-to-trade at any venue in any asset class" with no measurement.Source
    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
  • 02

    The homepage's second button, "Preview markets", opens the alpha trading app. When our logged-out reader loaded it without a live data connection, every stat showed "—", with "No trades in 24h", "Order book is empty" and "Couldn't load chart data". Crawlers, link previews and any visitor whose feed fails see the same thing, and it is a trader's first look at liquidity. The facts a visitor needs before signing up (three spot pairs, tick sizes, a 1 USDC minimum, zero fees and the rebate) sit in the docs instead.

    High impactLow effort First result in 2 weeks Co Founder & CTO

    First action

    Open "Preview markets" in a private window on a phone and record what shows in the first three seconds.

    How to measure it

    Signup clicks from the markets page

    Now
    –
    Target
    +30% vs the app link
    By
    60 days

    Steps

    1. 1

      Check the real first view

      Private window, phone and desktop, on a slow network. Record what renders.

      Co-Founder & CTO 1 hour

    2. 2

      Draft /markets

      Pair cards with specs, fees and rebates taken from the docs.

      Founding Software Engineer 2 days

    3. 3

      Add server-side stats

      Last price, 24h volume and spread, refreshed every minute, labelled alpha.

      Founding Software Engineer 3 days

    4. 4

      Fix the empty state

      Replace "Couldn't load chart data" with an explanation and a retry.

      Founding Software Engineer 1 day

    5. 5

      Repoint the button

      Homepage "Preview markets" to /markets.

      Founding Software Engineer 1 hour

    6. 6

      Measure

      Signup clicks from /markets compared with the old app link.

      Co-Founder & CEO 4 weeks

    First result
    2 weeks
    Cost
    Staff time only
    Owner
    Co Founder & CTO

    Smallest test

    Publish a static /markets specs table and repoint the button this week.

    Why it works

    Build pax.markets/markets as a server-rendered page with one card per pair (BTC-USDC, ETH-USDC, SOL-USDC). Each card shows tick size, minimum order, the free-tier fee and rebate, and the last price, 24h volume and spread from public market data, refreshed every minute and time-stamped. Point the homepage button there and keep a link to the live app for depth. In the app, swap the chart error for a short note that explains the alpha and offers a retry.

    Traders judge a venue on visible specs and activity before they start KYC. A page that loads everywhere is more honest than a blank book and converts better. It can also rank for searches like "zero fee BTC USDC".

    Risks and how to handle them

    • Thin volume becomes publicly visible

      Lead with specs and fees, and add volume once founding market makers quote (idea 11).

    • Stale numbers

      Time-stamp every figure and hide it if older than five minutes.

    Evidence

    • Your websiteThe "Preview markets" page (the trading app, labelled alpha), as our logged-out reader saw it without a live data connection: last price and 24h and 30d volume shown as "—", "No trades in 24h", "Order book is empty" and "Couldn't load chart data". The deposit panel offers USDC, with USD bank transfers "Coming soon".Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
    • DocsMarkets page: three spot markets, BTC-USDC, ETH-USDC and SOL-USDC, each with a 1 USDC minimum notional; no perpetual markets are listed.Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
  • 03

    The homepage compares PAX with a "Typical exchange fee −10 bps" and says "You keep 10.25 bps". A new trader at Kraken Pro's first tier pays "0.40%" maker and "0.80%" taker, which is 40 to 80 bps, so on the retail side PAX understates its own advantage by up to eight times. Basis points also mean little to most retail buyers.

    High impactLow effort First result in 3 weeks Co Founder & CEO

    First action

    Collect the current first-tier fees of three rival venues, with links and dates.

    How to measure it

    Homepage signup clicks

    Now
    –
    Target
    +20%
    By
    60 days

    Steps

    1. 1

      Gather fee tables

      First-tier maker/taker fees from rivals' own fee pages, dated.

      Co-Founder & CEO 1 day

    2. 2

      Pick defaults

      Default trade size ($1,000) and default comparison venue.

      Co-Founder & CEO 1 hour

    3. 3

      Build the calculator

      Size input, venue and tier picker, dollars first, bps toggle.

      Founding Software Engineer 2 days

    4. 4

      Compliance check

      Fair-comparison wording and sources reviewed.

      Chief Compliance Officer 1 day

    5. 5

      Ship and test

      A/B against the current bps example.

      Founding Software Engineer 3 weeks

    6. 6

      Refresh quarterly

      Re-check every rival's fee table.

      Co-Founder & CEO quarterly

    First result
    3 weeks
    Cost
    $0, staff time only
    Owner
    Co Founder & CEO

    Smallest test

    Change the homepage example to "$1,000 trade: up to $8 elsewhere, $0 here", with a source link.

    Why it works

    Replace the bps example with a small calculator. The trader enters a trade size and picks a venue and tier. The calculator shows the dollar cost elsewhere next to $0 plus the rebate at PAX, for example "$1,000 taker order: $8.00 at Kraken Pro tier 1, $0 at PAX". Use dated, sourced fee tables and keep a bps toggle for professionals. Be fair: include low-fee venues such as Hyperliquid (0.070% taker).

    Dollar amounts on a trader's own trade size are concrete, and showing the honest range builds trust with the free-tier traders whose order flow makes λ worth buying.

    Risks and how to handle them

    • Rival fees change

      Show the date and source under every number.

    • Looks cherry-picked

      Include the cheapest rivals, not just the most expensive.

    Evidence

    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
    • Competitor researchRivals' own pages: Kraken Pro's first-tier spot fees are "0.40%" maker and "0.80%" taker; Hyperliquid's spot base rate is 0.070% taker and 0.040% maker; EDX Markets runs "An institution-only spot venue", centrally cleared, and reports a $76 million Series C led by SBI Holdings.Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
  • 04

    The docs list New York and Tennessee as "not available" and Hawaii as business only. Abroad, the UK and Australia are business only, and 51 jurisdictions are prohibited, including Canada, India, Japan and Singapore. The homepage and signup path say nothing about eligibility, so these visitors find out only after they fill in forms or start verification.

    Medium impactLow effort First result in 2 weeks Chief Compliance Officer

    First action

    Count last month's signups that failed verification because of residency.

    How to measure it

    Signup-to-verified rate

    Now
    –
    Target
    +15 points
    By
    60 days

    Steps

    1. 1

      Size the problem

      Signups rejected for residency in the last 90 days.

      Chief Compliance Officer 1 day

    2. 2

      Write the eligibility page

      Plain-language version of the docs lists, kept in sync.

      Chief Compliance Officer 1 day

    3. 3

      Add the homepage line

      Under Sign up, linking to the page.

      Founding Software Engineer 1 hour

    4. 4

      Picker first

      Country and state as step one of signup.

      Founding Software Engineer 2 days

    5. 5

      Waitlist and business route

      Waitlist for excluded areas; business onboarding for business-only areas.

      Founding Software Engineer 2 days

    6. 6

      Track

      Verified rate and waitlist size by jurisdiction.

      Co-Founder & CEO monthly

    First result
    2 weeks
    Cost
    $0, staff time only
    Owner
    Chief Compliance Officer

    Smallest test

    Add the one-line eligibility note and link under Sign up today.

    Why it works

    Add one line under the homepage Sign up button ("Available in most US states and 140+ countries. See where") linking to a plain eligibility page. Put a country and state picker as the first signup step. Excluded visitors join a waitlist instead of a dead end, and business-only visitors in the UK, Australia and Hawaii go straight to business onboarding.

    It saves verification costs and support tickets on applicants who cannot be approved. It also turns rejected visitors into a list PAX can use when jurisdictions open, and it shows compliance discipline, which institutions look for.

    Risks and how to handle them

    • Fewer raw signups

      Measure verified and funded accounts, not raw signups.

    • Lists drift out of sync

      Generate the page and picker from one source.

    Evidence

    • DocsSupported Countries: 247 jurisdictions, of which 143 supported, 51 enhanced due diligence, 2 business only (Australia, United Kingdom) and 51 prohibited, including Canada, India, Japan and Singapore. Supported US States: New York and Tennessee "not available", Hawaii business only.Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
  • 05

    The app's deposit panel offers USDC, with USD bank transfers marked "Coming soon". Every market is quoted in USDC, and fees and rebates settle in USDC. A new trader who holds only dollars at a bank has no in-app path to a first trade, and nothing on the site explains one.

    High impactLow effort First result in 2 weeks Co Founder & CEO

    First action

    List the networks PAX accepts for USDC deposits and write the five-step guide.

    How to measure it

    Verified-to-funded rate

    Now
    –
    Target
    +20%
    By
    60 days

    Steps

    1. 1

      Map funding paths

      Supported USDC networks and the usual bank-to-USDC routes.

      Co-Founder & CTO 1 day

    2. 2

      Write the guide

      Five steps, timings, network warnings.

      Co-Founder & CEO 1 day

    3. 3

      Compliance review

      No endorsement of specific providers; clear risk notes.

      Chief Compliance Officer 1 day

    4. 4

      Place it in the app

      Deposit panel link and empty-balance prompt.

      Founding Software Engineer 1 day

    5. 5

      Email new accounts

      Guide sent after verification, with the USD waitlist checkbox.

      Co-Founder & CEO 1 day

    6. 6

      Track

      Time from verification to first deposit.

      Co-Founder & CEO weekly

    First result
    2 weeks
    Cost
    $0, staff time only
    Owner
    Co Founder & CEO

    Smallest test

    Add a single "How to get USDC" link to the deposit panel.

    Why it works

    Write a short guide, linked from the deposit panel, the post-signup email and the markets page. It covers how to get USDC from a bank account through common regulated on-ramps, which networks PAX accepts for USDC deposits, how long each step takes, and what a minimum first trade costs (1 USDC minimum). Add a "waiting on USD deposits? Get notified" checkbox so PAX can bring dollar-only users back when bank transfers launch.

    The gap between a verified account and a first trade is where most exchange signups stall, and a funding guide is the cheapest fix until USD rails ship.

    Risks and how to handle them

    • Users send USDC on the wrong network

      Bold network warning and a test-deposit suggestion.

    Evidence

    • Your websiteThe "Preview markets" page (the trading app, labelled alpha), as our logged-out reader saw it without a live data connection: last price and 24h and 30d volume shown as "—", "No trades in 24h", "Order book is empty" and "Couldn't load chart data". The deposit panel offers USDC, with USD bank transfers "Coming soon".Source
    • DocsMarkets page: three spot markets, BTC-USDC, ETH-USDC and SOL-USDC, each with a 1 USDC minimum notional; no perpetual markets are listed.Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
  • 06

    The homepage scores 95 on mobile and 100 on desktop, so these are small polish items. In the phone test, though, the "Trade In Nanoseconds" headline appears at 2.8 s against a 2.5 s target, held back by about 1.25 s of render-blocking Google Fonts. The viewport tag also blocks pinch zoom, and there is no social share image, so links shared in trader chats and on LinkedIn show a bare preview. A site that sells nanoseconds should not miss a speed target.

    Low impactLow effort First result in 1 day Founding Software Engineer

    First action

    Add the og:image and fix the viewport tag.

    How to measure it

    Mobile LCP

    Now
    –
    Target
    under 2.0 s
    By
    2 weeks

    Steps

    1. 1

      Share image

      1200x630 og:image and twitter:card tags.

      Founding Software Engineer 1 hour

    2. 2

      Viewport

      Remove user-scalable=no and the low maximum-scale.

      Founding Software Engineer 10 minutes

    3. 3

      Fonts

      Self-host, preload and swap.

      Founding Software Engineer 1 hour

    4. 4

      Link text

      Replace generic link labels.

      Founding Software Engineer 30 minutes

    5. 5

      Retest

      PageSpeed and a link-preview check.

      Founding Software Engineer 30 minutes

    6. 6

      Repeat on blog and docs

      Same share-image and viewport fixes.

      Founding Software Engineer 2 hours

    First result
    1 day
    Cost
    $0
    Owner
    Founding Software Engineer

    Smallest test

    Ship the share image today and paste the homepage link into a chat to check.

    Why it works

    Self-host and preload the web fonts (or use font-display: swap), remove the zoom-blocking viewport attributes, add an og:image (the latency number from idea 1 on a dark card works well), and give links descriptive text.

    Each fix takes minutes. The share card affects every link the team and early users post, and passing the phone LCP target matches the brand promise.

    Risks and how to handle them

    • Font flash changes the look

      Use size-adjusted fallbacks.

    Evidence

    • PageSpeedHomepage (November 9, 2026): mobile score 95, desktop 100 (desktop main content in 0.76 s). On the phone test the headline appears at 2.8 s against a 2.5 s target, with about 1.25 s of render-blocking requests (Google Fonts, 72 KB). The viewport tag blocks pinch zoom, the homepage has no social share image, and some links lack descriptive text.Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
  • 07

    PAX has a LinkedIn company page (listed on its YC profile) but no X account, and the site footer links no social profile at all. There is no newsletter, and the blog has been silent since February 2025. Every visitor who isn't ready to sign up today leaves with no way to hear from PAX again, even though the founders write well on market structure (the PFOF post).

    Medium impactLow effort First result in 1 month Co Founder & CEO

    First action

    Add the LinkedIn link and an email field to the footer.

    How to measure it

    Subscribers

    Now
    –
    Target
    1,000
    By
    6 months

    Steps

    1. 1

      Footer links

      LinkedIn and an email signup.

      Founding Software Engineer 1 hour

    2. 2

      Pick a tool

      A simple newsletter tool with double opt-in.

      Co-Founder & CEO 1 hour

    3. 3

      Write issue one

      What shipped, one argument, the numbers.

      Co-Founder & CEO half a day

    4. 4

      Compliance check

      Standard risk wording in the footer.

      Chief Compliance Officer 1 hour

    5. 5

      Cross-post

      LinkedIn company page and founder profiles.

      Co-Founder & CEO monthly

    6. 6

      Decide on X

      Test a handle for one quarter.

      Co-Founder & CEO 1 quarter

    First result
    1 month
    Cost
    Under $50 a month
    Owner
    Co Founder & CEO

    Smallest test

    Add the footer email field this week and see how many sign up before the first issue.

    Why it works

    Add LinkedIn and a newsletter signup to the footer and the blog. Send one short monthly note: what shipped (markets, λ modes, latency changes), one market-structure argument, and the month's fee and rebate numbers. Cross-post it to LinkedIn and decide whether an X account is worth running for crypto traders.

    Exchange decisions take months for market makers and weeks for traders. A monthly touch keeps PAX in mind until they're ready, and gives events and outreach somewhere to send people.

    Risks and how to handle them

    • Founders run out of time

      Keep it to 300 words; reuse the market-quality report (idea 12).

    Evidence

    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
    • Directory listingsYC company profile: "Co-located trading facility and matching engine on one chip", describing deterministic nanosecond tick-to-trade latency; W25, founded 2024, team of 6; it links a LinkedIn company page and no X account. A third-party product directory repeats that PAX "reduces exchange response latency by 1000x".Source
    • ResearchSite scan (November 9, 2026): no analytics or marketing tools detected on pax.markets, no newsletter and no testimonials; the headline changed from "The Future Of Crypto Exchange" to "Trade In Nanoseconds" since November 2025; no Hacker News mentions.Source
    • Your websiteBlog: four posts, from "In Crypto, Fees are High and Speeds are Slow" (November 6, 2024) to the PFOF post (February 2025), none since. The λ API post claims "the fastest tick-to-trade at any venue in any asset class" with no measurement.Source
  • 08

    Our scan detected no analytics or marketing tools on pax.markets. Visitors move from the marketing site to a separate app for signup, verification and deposit. Without a joined-up funnel, PAX cannot tell whether the latency message, the fee example or the markets preview drives signups, or where verified users stall before funding.

    Medium impactLow effort First result in 1 week Co Founder & CTO

    First action

    Choose a cookieless analytics tool and add it to the homepage.

    How to measure it

    Funnel coverage

    Now
    –
    Target
    5 of 5 events by source
    By
    30 days

    Steps

    1. 1

      Pick the tool

      Cookieless, no consent banner needed where that applies.

      Co-Founder & CTO 1 hour

    2. 2

      Tag the site and docs

      Page views and button clicks.

      Founding Software Engineer half a day

    3. 3

      Carry the source

      UTM and referrer through to signup.

      Founding Software Engineer 1 day

    4. 4

      App events

      Verified, first deposit, first trade, first λ program.

      Founding Software Engineer 1 day

    5. 5

      Privacy review

      Update the privacy and cookie policies.

      Chief Compliance Officer 1 day

    6. 6

      Weekly dashboard

      Conversion by step and source.

      Co-Founder & CEO weekly

    First result
    1 week
    Cost
    Under $50 a month
    Owner
    Co Founder & CTO

    Smallest test

    Tag the two homepage buttons and count clicks for one week.

    Why it works

    Add cookieless, privacy-first analytics to pax.markets, the docs and the app. Pass UTM and referrer through to signup, and record five events: signup started, verified, first deposit, first trade, first λ program. Build one weekly dashboard by source (events, launch post, blog, direct).

    With a team of six, every hour should go to the step that leaks most, and only a funnel shows which one that is. It also makes every other idea here measurable.

    Risks and how to handle them

    • Tracking in a regulated app

      Track events only, never balances or identity data.

    Evidence

    • ResearchSite scan (November 9, 2026): no analytics or marketing tools detected on pax.markets, no newsletter and no testimonials; the headline changed from "The Future Of Crypto Exchange" to "Trade In Nanoseconds" since November 2025; no Hacker News mentions.Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source

This quarter 11

The core plays for the next three months.

  • 09

    The docs promise a Binance-compatible API for traders moving from other venues, but the page says "This API is under construction. Documentation will be updated when available." Only three compatibility endpoints are listed. Many crypto trading bots already speak Binance's API, and those bots are the taker flow that makes λ worth buying for market makers: a web-app or standard-API order meeting a λ maker is "the most common scenario at PAX".

    High impactMedium effort First result in 6 weeks Founding Software Engineer

    First action

    List the ten Binance endpoints that common open-source bots call most and check which PAX already covers.

    How to measure it

    Active compatibility-API accounts

    Now
    –
    Target
    50
    By
    1 quarter

    Steps

    1. 1

      Pick the endpoint set

      The minimum bots need: order, cancel, open orders, balances, book ticker.

      Founding Software Engineer 2 days

    2. 2

      Build order entry

      Map to native PAX orders.

      Founding Software Engineer 3 weeks

    3. 3

      Build market data

      Top-of-book stream from the native feed.

      Founding Software Engineer 1 week

    4. 4

      Write the guide

      Base URL, symbols, differences, sandbox test.

      Founding Software Engineer 2 days

    5. 5

      Test with real bots

      Two popular open-source bots against the sandbox.

      Founding Software Engineer 1 week

    6. 6

      Announce

      Newsletter, LinkedIn and bot communities.

      Co-Founder & CEO 1 day

    7. 7

      Track usage

      Accounts, orders and errors by endpoint.

      Co-Founder & CTO weekly

    First result
    6 weeks
    Cost
    About six engineer-weeks
    Owner
    Founding Software Engineer

    Smallest test

    Make order submit and cancel compatible and invite five bot users to the sandbox.

    Why it works

    Ship the Binance-compatible order entry and top-of-book market data the page describes. Then publish a migration guide covering the base URL swap, supported and unsupported endpoints, symbol mapping (BTCUSDC to market 1), error-code differences and a ten-line test in the sandbox. Mark the page "beta" with a dated changelog instead of "under construction".

    Changing a base URL is a far smaller ask than rewriting a bot, and a zero-fee venue that pays rebates is a natural extra venue for existing bots to route to.

    Risks and how to handle them

    • Engineering time pulled from the core engine

      Ship the minimum endpoint set first; the docs already recommend the native API for advanced use.

    • Trademark use

      Describe compatibility factually and have compliance review the wording.

    Evidence

    • DocsBinance Compatible API page: "This API is under construction. Documentation will be updated when available." Only three compatibility endpoints are listed (order cancel, submit-batch and cancel-batch); Binance-compatible market data is described as planned.Source
    • Your websiteRebates post: a web-app or standard-API order meeting a λ maker is "the most common scenario at PAX"; when λ capacity is constrained, "λ fees rise and rebates rise with them".Source
    • DocsMarkets page: three spot markets, BTC-USDC, ETH-USDC and SOL-USDC, each with a 1 USDC minimum notional; no perpetual markets are listed.Source
  • 10

    A λ program is a new primitive: a JSON predicate on a market event, plus an order that fires on-chip. Only two placement modes exist today, and the docs describe no testing or simulation tool. The λ maker fee is 1/4 of the spread, up to 2 bps, and its value lies in faster cancels and queue priority. A market maker can't tell whether that trade-off pays without running their strategy against real event sequences, so the buying decision stalls at "interesting".

    High impactMedium effort First result in 1 quarter Founding Software Engineer

    First action

    Record one week of sandbox and production event streams in a replayable format.

    How to measure it

    Replay-to-live λ conversion

    Now
    –
    Target
    30% of firms that run a replay
    By
    2 quarters

    Steps

    1. 1

      Record events

      Store event streams in a replayable format.

      Founding Software Engineer 1 week

    2. 2

      Build the replay engine

      Software model of predicate matching, consistent with the chip.

      Founding Software Engineer 3 weeks

    3. 3

      Check against hardware

      Same inputs on the FPGA and the model must agree.

      Founding Hardware Engineer 1 week

    4. 4

      Report format

      Fires, fills, avoided adverse fills, fees and rebates.

      Co-Founder & CTO 3 days

    5. 5

      Docs and example

      Add a "replay this example" button to the existing cancel example.

      Founding Software Engineer 2 days

    6. 6

      Pilot with three makers

      Collect feedback on missing predicate types.

      Co-Founder & CEO 3 weeks

    First result
    1 quarter
    Cost
    About six engineer-weeks
    Owner
    Founding Software Engineer

    Smallest test

    Run the docs' cancel example against one recorded hour and email the report to two prospects.

    Why it works

    Add a replay mode to the self-serve sandbox. Upload a λ program, choose a recorded day of PAX (or synthetic) market events, and get a report: how often the predicate fired, fills, cancels that would have avoided adverse fills, λ fees paid and rebates generated for counterparties. Offer it as a JSON-in, JSON-out API plus a simple web view.

    Quants buy what they can backtest. A simulator turns a novel concept into a number in the buyer's own model, and it shows PAX which predicate types to build next.

    Risks and how to handle them

    • The model differs from the chip

      Run hardware-in-the-loop checks and publish known differences.

    • Replay data reveals other participants' activity

      Use anonymised or synthetic streams.

    Evidence

    • DocsWriting a λ program: a JSON predicate (market event) plus an order, submitted over REST; only two placement modes (ENTIRE and DELTA_QTY_RMD) are supported today and more are planned; no testing or simulation tool is described. The docs offer a self-serve sandbox with test identities, test funds and test API keys, and list no SDKs or code samples.Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
  • 11

    PAX's model only works when λ makers quote: they pay the fees that fund free-tier rebates, and a web-app order meeting a λ maker is "the most common scenario at PAX". Read without a live feed, the markets preview shows an empty book. PAX has announced λ priority tiers at 1.0 to 4.5 bps, so the first makers have reason to wait for terms.

    High impactMedium effort First result in 1 quarter Co Founder & CEO

    First action

    Draft a one-page term sheet and send it to ten firms from the CEO's trading network.

    How to measure it

    Two-sided quoting uptime

    Now
    –
    Target
    95% on BTC-USDC
    By
    1 quarter

    Steps

    1. 1

      Draft terms

      Rates, obligations, duration, exit clauses.

      Co-Founder & CEO 3 days

    2. 2

      Compliance review

      Fair access and disclosure of programme terms.

      Chief Compliance Officer 1 week

    3. 3

      Shortlist firms

      Twenty crypto market makers and prop shops.

      Co-Founder & CEO 2 days

    4. 4

      Pitch with evidence

      Latency page (idea 1) and replay results (idea 10).

      Co-Founder & CEO 4 weeks

    5. 5

      Onboard

      Sandbox, then production λ.

      Co-Founder & CTO 4 weeks

    6. 6

      Monitor obligations

      Daily quoting stats per maker.

      Co-Founder & CTO daily

    7. 7

      Publish the terms

      Public programme page.

      Co-Founder & CEO 1 day

    First result
    1 quarter
    Cost
    Forgone priority-tier revenue for the first year
    Owner
    Co Founder & CEO

    Smallest test

    Get one maker to commit to quoting BTC-USDC for 30 days.

    Why it works

    Offer up to five founding makers twelve months at base λ rates (no priority-tier pricing) and early access to new predicate types. In return they commit to minimum quoted size, maximum spread and uptime on BTC-USDC, ETH-USDC and SOL-USDC. Publish the programme terms on the site, and the anonymised aggregate results in the weekly market-quality report (idea 12).

    Committed quotes solve the cold start for takers, and founding terms give makers a reason to integrate now rather than after liquidity appears. Published terms also signal fairness to free-tier traders.

    Risks and how to handle them

    • Makers sign but don't quote

      Make founding terms conditional on meeting obligations each month.

    • Fair-access concerns

      Same published terms for any firm that meets the obligations.

    Builds on

    Evidence

    • Your websiteRebates post: a web-app or standard-API order meeting a λ maker is "the most common scenario at PAX"; when λ capacity is constrained, "λ fees rise and rebates rise with them".Source
    • Your websiteThe "Preview markets" page (the trading app, labelled alpha), as our logged-out reader saw it without a live data connection: last price and 24h and 30d volume shown as "—", "No trades in 24h", "Order book is empty" and "Couldn't load chart data". The deposit panel offers USDC, with USD bank transfers "Coming soon".Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
    • Your websiteTeam page: the CEO built a high-speed futures strategy on custom silicon at Citadel Securities and ran crypto HFT research; the CTO built low-latency crypto trading systems; the Chief Compliance Officer was Chief AML and Sanctions Officer at EDX Markets.Source
  • 12

    PAX's argument against payment for order flow is that price competition on an open exchange gives traders a better deal, but nothing public shows the market's quality. Read without a live feed, the app shows no stats, and the rebate is described only as "up to" 0.25 bps, depending on order type.

    Medium impactMedium effort First result in 1 month Co Founder & CTO

    First action

    Define the six metrics and compute them for last week.

    How to measure it

    Signups attributed to the report

    Now
    –
    Target
    50 a month
    By
    1 quarter

    Steps

    1. 1

      Define metrics

      Six numbers with exact formulas.

      Co-Founder & CTO 1 day

    2. 2

      Automate

      Weekly job from the market data stream.

      Founding Software Engineer 1 week

    3. 3

      Page template

      One page per week, with an archive.

      Founding Software Engineer 2 days

    4. 4

      Compliance review

      Wording and no forward-looking claims.

      Chief Compliance Officer 1 day

    5. 5

      Distribute

      Newsletter, LinkedIn and maker outreach.

      Co-Founder & CEO weekly

    6. 6

      Review

      Which metric moves signups.

      Co-Founder & CEO monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Co Founder & CTO

    Smallest test

    Publish one week's numbers as a blog post.

    Why it works

    Every Monday, publish a page and an email covering each pair's time-weighted spread, depth at 10 bps, two-sided uptime, matching engine uptime, total rebates paid to free-tier traders and the share of fills that earned a rebate. Keep an archive so the trend is visible.

    Takers trust numbers over adjectives. Makers can see the competition they would face, and the report gives the newsletter, LinkedIn and sales a fresh proof point every week.

    Risks and how to handle them

    • Early numbers look weak

      Start once founding makers quote (idea 11), and report the trend, not just the level.

    Builds on

    Evidence

    • Your websiteThe "Preview markets" page (the trading app, labelled alpha), as our logged-out reader saw it without a live data connection: last price and 24h and 30d volume shown as "—", "No trades in 24h", "Order book is empty" and "Couldn't load chart data". The deposit panel offers USDC, with USD bank transfers "Coming soon".Source
    • Your websiteRebates post: a web-app or standard-API order meeting a λ maker is "the most common scenario at PAX"; when λ capacity is constrained, "λ fees rise and rebates rise with them".Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
  • 13

    Every fill already reports a fee_rebate value, settled in USDC, but the rebate is small per trade (up to 0.25 bps) and nothing on the site or in the logged-out app promises a running total of what a trader has earned. The real benefit, paying no fee where other venues charge 40 to 80 bps, is invisible after signup.

    Medium impactLow effort First result in 1 month Founding Software Engineer

    First action

    Sum fee_rebate per account for last month.

    How to measure it

    Monthly retention of funded traders

    Now
    –
    Target
    +10 points
    By
    1 quarter

    Steps

    1. 1

      Aggregate rebates

      Per account, lifetime and monthly.

      Founding Software Engineer 2 days

    2. 2

      Fees-avoided estimate

      Reference fee tables, dated.

      Co-Founder & CEO 1 day

    3. 3

      Design the card

      Dollars first, bps on hover.

      Founding Software Engineer 2 days

    4. 4

      Compliance wording

      Make clear the comparison is an estimate.

      Chief Compliance Officer 1 day

    5. 5

      Monthly statement email

      Savings, trades and new markets.

      Founding Software Engineer 2 days

    6. 6

      Measure

      Retention of traders who saw the card.

      Co-Founder & CEO monthly

    First result
    1 month
    Cost
    Staff time only
    Owner
    Founding Software Engineer

    Smallest test

    Email ten active traders their savings for last month and see how many reply or trade more.

    Why it works

    Add a "Your savings" card to the account page showing lifetime and monthly cash-back from fee_rebate, plus estimated fees avoided against a reference venue the trader picks (same tables as idea 3). Include it in a monthly statement email.

    Visible, growing savings reinforce the reason a trader chose PAX and give them a number to share.

    Risks and how to handle them

    • Tiny rebate numbers look trivial

      Lead with fees avoided; show cash-back as the extra.

    Builds on

    Evidence

    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
    • Competitor researchRivals' own pages: Kraken Pro's first-tier spot fees are "0.40%" maker and "0.80%" taker; Hyperliquid's spot base rate is 0.070% taker and 0.040% maker; EDX Markets runs "An institution-only spot venue", centrally cleared, and reports a $76 million Series C led by SBI Holdings.Source
  • 14

    The docs cover REST, WebSocket and λ endpoints in detail but list no SDKs or code samples. A new integrator has to hand-build authentication, book building and λ JSON before placing a test order. The team already writes Rust, per its hiring posts.

    Medium impactMedium effort First result in 1 month Founding Software Engineer

    First action

    Write the sandbox first-order example in Python, in under 30 lines.

    How to measure it

    Time to first API order

    Now
    –
    Target
    under 30 minutes
    By
    1 quarter

    Steps

    1. 1

      Scope

      Auth, orders, book, λ.

      Founding Software Engineer 1 day

    2. 2

      Python client

      With type hints and tests.

      Founding Software Engineer 1 week

    3. 3

      Rust client

      Async, for latency-sensitive users.

      Founding Software Engineer 1 week

    4. 4

      Three examples

      Sandbox order, local book, λ cancel.

      Founding Software Engineer 3 days

    5. 5

      Publish

      Package registries and a public repository.

      Co-Founder & CTO 1 day

    6. 6

      Link from docs

      Header and Get Started.

      Founding Software Engineer 1 hour

    First result
    1 month
    Cost
    About three engineer-weeks
    Owner
    Founding Software Engineer

    Smallest test

    Add one Python snippet to the Sandbox docs page.

    Why it works

    Publish thin, open-source Python and Rust clients covering auth, order submit and cancel, the WebSocket book builder and λ submit, with three runnable examples: a sandbox order, a local order book and the docs' λ cancel example. Link them from every docs page header.

    Time to first order is the activation metric for API traders. Libraries cut it from days to minutes and make PAX easy to add as an extra venue.

    Risks and how to handle them

    • Maintenance load

      Generate clients from the API spec where possible.

    Evidence

    • DocsWriting a λ program: a JSON predicate (market event) plus an order, submitted over REST; only two placement modes (ENTIRE and DELTA_QTY_RMD) are supported today and more are planned; no testing or simulation tool is described. The docs offer a self-serve sandbox with test identities, test funds and test API keys, and list no SDKs or code samples.Source
    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
  • 15

    PAX's custody and licensing story sits in a hero-line disclosure and a footer: regulated services and custody are provided by 1Money USA, Inc., a licensed money transmitter (NMLS ID 2628653); PAX "is not a bank"; holdings are not FDIC or SIPC protected. The Chief Compliance Officer was Chief AML and Sanctions Officer at EDX Markets, an institution-only, centrally cleared rival, yet the site never presents that compliance depth as a reason to trust PAX. Institutions and cautious retail traders look for exactly this before depositing.

    Medium impactLow effort First result in 1 month Chief Compliance Officer

    First action

    Confirm with 1Money USA what PAX may say publicly about custody and segregation.

    How to measure it

    7-day funding rate

    Now
    –
    Target
    +10%
    By
    1 quarter

    Steps

    1. 1

      Agree the facts

      With the custody and licensing partner.

      Chief Compliance Officer 1 week

    2. 2

      Write /trust

      Five sections, plain language.

      Chief Compliance Officer 3 days

    3. 3

      Legal review

      No implied guarantees.

      Chief Compliance Officer 3 days

    4. 4

      Build and link

      Homepage, deposit panel, business onboarding.

      Founding Software Engineer 1 day

    5. 5

      Business pack

      A PDF version for institutional due diligence.

      Chief Compliance Officer 1 week

    6. 6

      Review quarterly

      Keep licences and partners current.

      Chief Compliance Officer quarterly

    First result
    1 month
    Cost
    Staff time; possible legal review fees
    Owner
    Chief Compliance Officer

    Smallest test

    Add a three-line "Where your money is held" box to the deposit panel.

    Why it works

    Publish /trust with five short sections: who PAX is and who 1Money USA is, with a link to its licences; where customer funds and crypto sit and how they are segregated; what is and is not insured; the AML and sanctions programme and who leads it; and status and incident history. Link it from the homepage, the deposit panel and the business onboarding flow.

    A new venue's biggest objection is "is my money safe?" A clear, factual answer near the deposit button removes it without overpromising.

    Risks and how to handle them

    • Saying more than the partner allows

      Use only partner-approved wording.

    Evidence

    • Your websiteHomepage disclosures: regulated services are provided by 1Money USA, Inc., a licensed money transmitter (NMLS ID 2628653), which also manages custody; PAX Markets Technology, LLC "is not a bank" and crypto holdings are not FDIC or SIPC protected; the page promises institutions third-party custody and settlement.Source
    • Your websiteTeam page: the CEO built a high-speed futures strategy on custom silicon at Citadel Securities and ran crypto HFT research; the CTO built low-latency crypto trading systems; the Chief Compliance Officer was Chief AML and Sanctions Officer at EDX Markets.Source
    • Competitor researchRivals' own pages: Kraken Pro's first-tier spot fees are "0.40%" maker and "0.80%" taker; Hyperliquid's spot base rate is 0.070% taker and 0.040% maker; EDX Markets runs "An institution-only spot venue", centrally cleared, and reports a $76 million Series C led by SBI Holdings.Source
  • 16

    PAX's revenue comes from λ fees paid mainly by market makers: 1/4 of the spread up to 2 bps, with priority tiers to come. The people who know which trading firms are looking for a new venue are prime brokers, crypto trading consultants, FPGA and low-latency vendors and former colleagues, and the founders come from that world (Citadel Securities, crypto HFT). Nothing on the site rewards an introduction.

    High impactMedium effort First result in 1 quarter Co Founder & CEO

    First action

    Ask the compliance officer which introducer arrangements are permitted and what disclosures they require.

    How to measure it

    Introduced firms live on λ

    Now
    –
    Target
    5
    By
    2 quarters

    Steps

    1. 1

      Regulatory check

      Permitted structure and required disclosures.

      Chief Compliance Officer 2 weeks

    2. 2

      Set terms

      Share, duration, cap, payment in USDC.

      Co-Founder & CEO 2 days

    3. 3

      Registration form

      Logs introductions with a timestamp.

      Founding Software Engineer 1 day

    4. 4

      Explainer

      One page an introducer can forward.

      Co-Founder & CEO 1 day

    5. 5

      Seed 20 introducers

      Founders' trading and silicon networks.

      Co-Founder & CEO 4 weeks

    6. 6

      Track and pay

      Monthly fee-share statements.

      Co-Founder & CTO monthly

    First result
    1 quarter
    Cost
    A share of new λ revenue only
    Owner
    Co Founder & CEO

    Smallest test

    Offer the terms to five people in the founders' network and count introductions in 30 days.

    Why it works

    Launch a written introducer programme: anyone who brings a firm that goes live on λ receives a share (for example 10%) of that firm's λ fees for twelve months, capped and paid in USDC. Publish a short programme page, a one-page explainer for introducers to forward, and a simple form that registers the introduction before first contact.

    Institutional sales run on warm introductions. Paying only from revenue earned keeps the cost aligned, and a six-person team gets a sales force without hiring one.

    Risks and how to handle them

    • Regulatory limits on paid referrals

      Have compliance sign off on the structure before any outreach.

    • Disputes over who introduced whom

      Register introductions first, then pay.

    Evidence

    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
    • Your websiteTeam page: the CEO built a high-speed futures strategy on custom silicon at Citadel Securities and ran crypto HFT research; the CTO built low-latency crypto trading systems; the Chief Compliance Officer was Chief AML and Sanctions Officer at EDX Markets.Source
  • 17

    For a taker, PAX costs nothing and pays up to 0.25 bps back, while first-tier fees elsewhere run from 7 bps (Hyperliquid taker) to 80 bps (Kraken Pro taker). A router that compares all-in cost would send marketable flow to PAX whenever its price is at or near the best. That flow is exactly what λ makers pay to trade against.

    Medium impactMedium effort First result in 1 quarter Co Founder & CEO

    First action

    List 15 routers and terminals and find which already support a new venue through a standard connector.

    How to measure it

    Partner-routed volume share

    Now
    –
    Target
    25%
    By
    2 quarters

    Steps

    1. 1

      Shortlist

      Fifteen routers and terminals with US-eligible users.

      Co-Founder & CEO 1 week

    2. 2

      Integration kit

      Docs, sandbox, fee data feed.

      Founding Software Engineer 2 weeks

    3. 3

      Outreach

      Founder-led calls.

      Co-Founder & CEO 4 weeks

    4. 4

      Support the build

      Shared channel per partner.

      Founding Software Engineer ongoing

    5. 5

      Compliance checks

      Partner users must meet PAX eligibility.

      Chief Compliance Officer per partner

    6. 6

      Launch

      Joint announcement and newsletter.

      Co-Founder & CEO 1 day

    First result
    1 quarter
    Cost
    Staff time; possible integration incentives
    Owner
    Co Founder & CEO

    Smallest test

    Get one terminal to add PAX as a read-only market data venue.

    Build the prospect list. The list compiled from public sources, verified, deduplicated and delivered to your CRM. Usually live within days.

    Book a working session

    Why it works

    Build a shortlist of crypto smart order routers, multi-venue terminals and portfolio trading tools used by funds and active traders. Offer each one an integration kit (native API, the Binance-compatible layer from idea 9, sandbox accounts and fee/rebate data in a machine-readable form) plus co-marketing at launch.

    One routing integration brings many traders at once, and a zero-fee, rebate-paying venue looks better to every algorithm that counts fees.

    Risks and how to handle them

    • Routers won't integrate a thin venue

      Lead with the founding market makers' quoting commitments (idea 11).

    Builds on

    Evidence

    • DocsFee Structure: the free tier (the default) pays no trading fee and earns half of a λ counterparty's fee as a rebate, "capped at a maximum of 0.25 bps"; λ makers pay 1/4 of the spread up to 2 bps and λ takers 0.5 bps; PAX "will introduce" λ priority tiers at 1.0, 1.5, 3.0 and 4.5 bps; fees and rebates settle in USDC and every fill reports a fee_rebate value.Source
    • Competitor researchRivals' own pages: Kraken Pro's first-tier spot fees are "0.40%" maker and "0.80%" taker; Hyperliquid's spot base rate is 0.070% taker and 0.040% maker; EDX Markets runs "An institution-only spot venue", centrally cleared, and reports a $76 million Series C led by SBI Holdings.Source
    • DocsBinance Compatible API page: "This API is under construction. Documentation will be updated when available." Only three compatibility endpoints are listed (order cancel, submit-batch and cancel-batch); Binance-compatible market data is described as planned.Source
  • 18

    PAX has had booths at Token2049 Singapore (October 2026), Token2049 Dubai and Consensus Miami. Its docs list Singapore, Canada, India and Japan among the 51 prohibited jurisdictions and the UK and Australia as business only. Part of every booth audience in Asia cannot become a customer. The UAE and the US are open, and the Events page lists nothing upcoming.

    Medium impactLow effort First result in next event Co Founder & CEO

    First action

    Tag the October Singapore lead list by country and count how many are eligible.

    How to measure it

    Event cost per funded account

    Now
    –
    Target
    -40%
    By
    next 3 events

    Steps

    1. 1

      Audit past leads

      Tag by country and type.

      Co-Founder & CEO 2 days

    2. 2

      Scoring sheet

      Eligible-audience share, firm share, cost.

      Co-Founder & CEO 1 day

    3. 3

      Pick the 2027 calendar

      Highest-scoring events first.

      Co-Founder & CEO 1 day

    4. 4

      Booth form

      Country first, then trader or firm.

      Founding Software Engineer 1 day

    5. 5

      Three follow-up tracks

      Trader, firm, waitlist.

      Co-Founder & CEO 2 days

    6. 6

      List upcoming events

      With a meeting-booking link.

      Founding Software Engineer 1 hour

    First result
    next event
    Cost
    $0 to plan; saves event budget
    Owner
    Co Founder & CEO

    Smallest test

    Re-tag the last event's leads and send the three tracks.

    Why it works

    Score each candidate 2027 event on the share of attendees from supported jurisdictions and the share of market makers and trading firms. At the booth, scan badges into a form whose first field is country, and send three follow-up tracks: eligible traders (signup plus the USDC guide), firms (λ, the latency page and the founding maker programme) and ineligible visitors (waitlist only). Publish upcoming events on the Events page so prospects can book meetings.

    Booths are a small team's largest marketing cost, and qualifying by eligibility puts that spend and follow-up time on people who can actually become customers.

    Risks and how to handle them

    • Missing global firms at excluded-country events

      Firms can be served as business clients where allowed; qualify by entity, not just residence.

    Builds on

    Evidence

    • Your websiteEvents page, titled "Past Events": booths at Token2049 Singapore (October 7–8, 2026), Consensus Miami (May 2026) and Token2049 Dubai (April 2026); sponsorships of Stanford's BASS summit (July 2026) and the Security Traders Association conference (October 2025); no upcoming events.Source
    • DocsSupported Countries: 247 jurisdictions, of which 143 supported, 51 enhanced due diligence, 2 business only (Australia, United Kingdom) and 51 prohibited, including Canada, India, Japan and Singapore. Supported US States: New York and Tennessee "not available", Hawaii business only.Source
  • 19

    The blog has four posts, the last from February 2025, and our scan found no Hacker News mentions. PAX is hiring a Founding Principal Hardware Engineer, a Founding Hardware Engineer and a Founding Software Engineer to build a Rust and SystemVerilog matching engine on AWS F2 FPGAs, and later an ASIC. The engineering story that would attract those hires is the same one that convinces quant buyers, and it is not being told.

    Medium impactMedium effort First result in 1 month Co Founder & CTO

    First action

    Draft "How a λ predicate fires in hardware" from the docs' cancel example.

    How to measure it

    Applicants and sandbox signups attributed to posts

    Now
    –
    Target
    20 applicants, 30 sandbox signups
    By
    2 quarters

    Steps

    1. 1

      Topic list

      Six posts, one a month.

      Co-Founder & CTO 1 day

    2. 2

      First post

      λ predicate in hardware.

      Co-Founder & CTO 1 week

    3. 3

      Review

      No sensitive implementation detail; compliance check.

      Chief Compliance Officer 2 days

    4. 4

      Add CTAs

      Open roles and sandbox link.

      Founding Software Engineer 1 hour

    5. 5

      Distribute

      Hacker News, FPGA and HFT forums, LinkedIn.

      Co-Founder & CEO per post

    6. 6

      Measure

      Applicants and sandbox signups by post.

      Co-Founder & CEO monthly

    First result
    1 month
    Cost
    About two days of engineering time a month
    Owner
    Co Founder & CTO

    Smallest test

    Publish one post and track applicants for two weeks.

    Why it works

    Publish one deep technical post a month, written by the CTO or a founding engineer. Topics: how the λ predicate is matched in hardware, what F2 FPGAs make possible, deterministic latency versus average latency, book building over WebSocket, and lessons from moving toward an ASIC. End each post with the open roles and a sandbox link, and submit the best ones to Hacker News and FPGA and HFT communities.

    Hard engineering writing attracts both the engineers PAX needs and the quant firms it sells to, and it is content competitors cannot copy.

    Risks and how to handle them

    • Revealing proprietary design

      Explain concepts and results, not implementation; the λ API is already patented.

    Evidence

    • Your websiteBlog: four posts, from "In Crypto, Fees are High and Speeds are Slow" (November 6, 2024) to the PFOF post (February 2025), none since. The λ API post claims "the fastest tick-to-trade at any venue in any asset class" with no measurement.Source
    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
    • ResearchSite scan (November 9, 2026): no analytics or marketing tools detected on pax.markets, no newsletter and no testimonials; the headline changed from "The Future Of Crypto Exchange" to "Trade In Nanoseconds" since November 2025; no Hacker News mentions.Source

Big bets 2

Larger plays that can change the trajectory.

  • 20

    PAX's launch post names NYSE, NASDAQ, Deutsche Börse, EDX, Crossover, Binance and OKX as slower venues, and its YC profile promises deterministic nanosecond tick-to-trade latency. Every claim so far is self-reported. Market makers and, later, equities broker-dealers will ask for third-party evidence before moving serious flow, and an audit is the answer a startup can't fake.

    High impactHigh effort First result in 2 quarters Co Founder & CEO

    First action

    Write the test protocol and ask two potential auditors for quotes.

    How to measure it

    Firms going live on λ after reviewing the audit

    Now
    –
    Target
    5
    By
    2 quarters after publication

    Steps

    1. 1

      Protocol

      Definition, tooling and load conditions, from idea 1.

      Co-Founder & CTO 2 weeks

    2. 2

      Select auditor

      Independent lab or academic group.

      Co-Founder & CEO 1 month

    3. 3

      Prepare the environment

      Production-equivalent build and access.

      Founding Principal Hardware Engineer 2 weeks

    4. 4

      Run the audit

      Including tail percentiles and jitter.

      Co-Founder & CTO 1 month

    5. 5

      Publish

      Report, data and an explainer post.

      Co-Founder & CEO 1 week

    6. 6

      Use in sales

      Attach to every institutional pitch.

      Co-Founder & CEO ongoing

    7. 7

      Repeat on ASIC

      Same protocol, new silicon.

      Founding Principal Hardware Engineer at ASIC launch

    First result
    2 quarters
    Cost
    Audit fees, likely tens of thousands of dollars
    Owner
    Co Founder & CEO

    Smallest test

    Invite one academic group to reproduce the sandbox measurement informally.

    Why it works

    Once the /latency method is published (idea 1), engage an independent testing lab or university research group to measure the round trip on the production build under a published protocol. Publish the full report, the raw percentile data and the test date, and repeat when the ASIC ships.

    An independent number turns the headline claim into due-diligence material, shortens institutional sales cycles and gives the press something verifiable to report.

    Risks and how to handle them

    • Audited number is worse than the marketing number

      Publish anyway and update the site; credibility is the asset.

    • Comparisons to other venues draw objections

      Audit PAX only; cite others' published figures separately.

    Builds on

    Evidence

    • Launch postLaunch YC post (listed as over a year old): "The fastest exchanges today boast a 30 microsecond round-trip latency. PAX crunches that to 30 nanoseconds (1000x faster)"; it claims PAX is faster than NYSE, NASDAQ, Deutsche Börse, EDX, Crossover, Binance and OKX, and describes the model as selling speed and buying order flow.Source
    • Directory listingsYC company profile: "Co-located trading facility and matching engine on one chip", describing deterministic nanosecond tick-to-trade latency; W25, founded 2024, team of 6; it links a LinkedIn company page and no X account. A third-party product directory repeats that PAX "reduces exchange response latency by 1000x".Source
    • Your websiteHomepage: H1 "Trade In Nanoseconds"; "That round trip takes about 20 µs at the fastest traditional venues and 1 to 10 ms at crypto venues"; "On the chip, the same round trip takes about 40 ns, 500x faster than the 20 µs"; fee example "Typical exchange fee −10 bps", "PAX rebate +0.25 bps", "You keep 10.25 bps"; buttons "Sign up" and "Preview markets"; revenue comes from selling λ API access, mainly to market makers; the footer links Team, Blog, Events, Careers, Contact, Docs and Status but no social profiles, and the page does not say where the service is available.Source
  • 21

    PAX's hiring posts name a U.S. equities ATS as the next step after crypto, and the CEO's background is high-speed trading at Citadel Securities. PAX has already sponsored the Security Traders Association conference, and its PFOF post argues for on-exchange incentives as an alternative to payment for order flow. Yet pax.markets says nothing about equities, so the securities-trading contacts PAX has started to build have nothing to sign up for.

    High impactHigh effort First result in 1 quarter Co Founder & CEO

    First action

    Have the compliance officer approve wording that describes the ATS as planned, with no offer of services.

    How to measure it

    Broker-dealers and trading firms on the waitlist

    Now
    –
    Target
    40
    By
    2 quarters

    Steps

    1. 1

      Compliance wording

      Planned, not offered; no solicitation.

      Chief Compliance Officer 2 weeks

    2. 2

      Write the design note

      On-chip logic, rebates, fair access.

      Co-Founder & CEO 2 weeks

    3. 3

      Waitlist page

      Firm type, role, interests.

      Founding Software Engineer 2 days

    4. 4

      Seed from events

      Securities Traders Association contacts and the CEO's network.

      Co-Founder & CEO 4 weeks

    5. 5

      Design-partner calls

      Ten firms; record must-have features.

      Co-Founder & CEO 6 weeks

    6. 6

      Decide

      Go or wait, based on demand and readiness.

      Co-Founder & CEO end of quarter 2

    First result
    1 quarter
    Cost
    Staff time; legal counsel review
    Owner
    Co Founder & CEO

    Smallest test

    Add a one-line "Interested in equities? Join the list" link to the Team page.

    Why it works

    Publish a short design note on how on-chip order logic and rebates funded by speed could work in an equities ATS, clearly marked as planned and subject to regulatory approval. Add a waitlist for broker-dealers and trading firms that asks for role, firm type and interest (λ, routing, data). Use the list to size demand, pick design partners and support fundraising.

    Equities subscribers take long to qualify. Starting the list now turns conference contacts into a measurable pipeline before the product exists, and tells PAX which features matter first.

    Risks and how to handle them

    • Regulatory sensitivity around pre-launch marketing

      Compliance-approved wording and no offer of services.

    • Distraction from crypto liquidity

      Cap effort at a page and ten calls until crypto metrics are met.

    Builds on

    Evidence

    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
    • Your websiteTeam page: the CEO built a high-speed futures strategy on custom silicon at Citadel Securities and ran crypto HFT research; the CTO built low-latency crypto trading systems; the Chief Compliance Officer was Chief AML and Sanctions Officer at EDX Markets.Source
    • Your websiteEvents page, titled "Past Events": booths at Token2049 Singapore (October 7–8, 2026), Consensus Miami (May 2026) and Token2049 Dubai (April 2026); sponsorships of Stanford's BASS summit (July 2026) and the Security Traders Association conference (October 2025); no upcoming events.Source

Wildcards 2

Unexpected experiments, each with a cheap way to test it.

  • 22

    PAX sponsored Stanford's BASS summit, its Founding Software and Founding Hardware Engineer roles welcome new graduates, and the self-serve sandbox already provides test identities, test funds and API keys. A λ program is a new way to express a trading strategy, and the students who learn it now are the next generation of quant hires at the firms PAX wants as customers.

    Medium impactMedium effort First result in 1 quarter Co Founder & CTO

    First action

    Ask the BASS organisers and two university quant clubs whether they would co-host.

    How to measure it

    Active challenge teams

    Now
    –
    Target
    40
    By
    1 quarter

    Steps

    1. 1

      Design the scoring

      PnL after fees, risk, predicate use.

      Co-Founder & CTO 1 week

    2. 2

      Prepare market data

      Recorded or simulated streams (idea 10).

      Founding Software Engineer 2 weeks

    3. 3

      Recruit clubs

      Three universities to start.

      Co-Founder & CEO 3 weeks

    4. 4

      Rules and eligibility

      Sandbox only, no real money, contest rules reviewed.

      Chief Compliance Officer 1 week

    5. 5

      Run it

      Four weeks with a weekly leaderboard.

      Co-Founder & CTO 4 weeks

    6. 6

      Hire and publish

      Interviews for top teams; results post.

      Co-Founder & CTO 2 weeks

    First result
    1 quarter
    Cost
    Prize pool of a few thousand dollars, plus staff time
    Owner
    Co Founder & CTO

    Smallest test

    Run a one-weekend challenge with a single university club.

    Why it works

    Run a four-week sandbox challenge for university teams. Each team writes λ programs against a recorded or simulated market, scored on PnL after λ fees, risk and the creativity of predicates used. Give the top teams cash prizes, interviews for PAX's founding roles and a write-up on the blog. Partner with two or three university finance, computer science or FPGA clubs to recruit teams.

    It fills the hiring pipeline, stress-tests the λ developer experience with fresh users, and seeds future customers inside trading firms, all for the price of the prize pool.

    Risks and how to handle them

    • Low participation

      Co-host with clubs that already run trading competitions.

    Builds on

    Evidence

    • Your websiteEvents page, titled "Past Events": booths at Token2049 Singapore (October 7–8, 2026), Consensus Miami (May 2026) and Token2049 Dubai (April 2026); sponsorships of Stanford's BASS summit (July 2026) and the Security Traders Association conference (October 2025); no upcoming events.Source
    • DocsWriting a λ program: a JSON predicate (market event) plus an order, submitted over REST; only two placement modes (ENTIRE and DELTA_QTY_RMD) are supported today and more are planned; no testing or simulation tool is described. The docs offer a self-serve sandbox with test identities, test funds and test API keys, and list no SDKs or code samples.Source
    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
  • 23

    PAX's hiring posts describe an ASIC that combines network aggregation with exchange functions, and a plan to eventually run exchanges in every asset class and in global markets. The launch post says the chip runs an exchange that is "faster and smaller than a datacenter". Running every venue itself means winning liquidity market by market, which is slow and capital-heavy. Many existing venues in crypto, prediction markets and regional securities markets would value the matching technology without PAX having to win their liquidity.

    High impactHigh effort First result in 2 quarters Co Founder & CEO

    First action

    List twenty venues that run their own matching engine and could benefit, and book five discovery calls.

    How to measure it

    Licensing LOIs

    Now
    –
    Target
    2
    By
    1 year

    Steps

    1. 1

      Map candidates

      Twenty venues with in-house engines.

      Co-Founder & CEO 2 weeks

    2. 2

      Discovery calls

      Pain points, budgets, regulatory constraints.

      Co-Founder & CEO 6 weeks

    3. 3

      Technical fit

      What would need to change in the engine.

      Founding Principal Hardware Engineer 4 weeks

    4. 4

      IP and terms

      Patent position and licence model.

      Co-Founder & CEO 4 weeks

    5. 5

      Go or no-go

      Compare against focus on PAX's own venue.

      Co-Founder & CEO 1 day

    6. 6

      Pilot

      One venue on the FPGA build.

      Co-Founder & CTO 2 quarters

    First result
    2 quarters
    Cost
    Founder time for discovery; pilot costs later
    Owner
    Co Founder & CEO

    Smallest test

    Five discovery calls with venue CTOs in one month.

    Why it works

    Quietly test a second business line: a licensed matching engine on PAX silicon, with or without the λ facility, for venues that want lower latency and smaller infrastructure. Start with ten discovery conversations, then decide whether a pilot with one venue (run on PAX's FPGA build) is worth the distraction. Keep PAX's own exchange as the flagship reference customer.

    Licensing earns revenue from the core technology without the cold start of a two-sided market, and each licensee validates the silicon for PAX's own venues.

    Risks and how to handle them

    • Arms rivals with PAX's edge

      License only to venues outside PAX's own markets, or without λ.

    • Distraction for a small team

      Discovery only until crypto liquidity targets are met.

    Evidence

    • HiringThree open founding roles on the YC job board, all in Redwood City with visa sponsorship: Founding Principal Hardware Engineer ($200K–$225K, 0.50%–1.00%), Founding Software Engineer and Founding Hardware Engineer ($150K–$200K, new grads welcome). The work: an ultra-low-latency matching engine built as a hardware/software co-design (Rust, SystemVerilog); launch on AWS F2 FPGA instances, then an ASIC combining network aggregation with exchange functions; crypto first with spot and perpetual futures for BTC and ETH, then a U.S. equities ATS, then other asset classes and markets.Source
    • Launch postLaunch YC post (listed as over a year old): "The fastest exchanges today boast a 30 microsecond round-trip latency. PAX crunches that to 30 nanoseconds (1000x faster)"; it claims PAX is faster than NYSE, NASDAQ, Deutsche Börse, EDX, Crossover, Binance and OKX, and describes the model as selling speed and buying order flow.Source

03

How Programmable can help

Where an idea in this playbook is something Programmable builds or runs, it's listed here with what we'd do. Everything else is yours to run.

Have it built

ProgrammableAgency

1 of the 23 ideas are things Programmable builds and runs for its customers: pages, outreach, prospect lists, booking flows and monitors. Pick one, or send the whole list.

Prospect lists1The list compiled from public sources, verified, deduplicated and delivered to your CRM.

A 30-minute session with the engineer who would build it, not a sales call. You leave with a plan whether or not you hire us.

Book a working session

Usually live within days.

Tools for the rest

ProgrammableTalk

Fill the events in this plan

These ideas are events. Talk gets registrants for them and tells you who's coming and what their company does before the day.

Plan an event (opens programmable.talk in a new tab)
More example playbooks and reviews

More growth playbooks

Other companies we have researched, each read from what they are hiring for.

  • PlayStation (Sony Interactive Entertainment) growth strategyMakes PS5 and PS5 Pro consoles, PS VR2 and PS Portal, publishes first-party games, and runs PlayStation Network, PlayStation Plus and the…21 growth ideas
  • Miles Ahead Hosting growth strategyWeb hosting, domain registration, custom website design and SEO, with AI automation and email marketing, for small businesses.21 growth ideas
  • MagicSchool growth strategyAn AI platform for schools: 80+ teacher tools for planning, assessment, feedback and paperwork, teacher-guided student tools, and an…22 growth ideas
  • Foundation Risk Partners growth strategyA national insurance brokerage and consulting firm built from close to 200 agency partnerships: commercial property and casualty, employee…21 growth ideas
  • Stable growth strategyAn AI-powered business address and virtual mailbox: permanent street addresses in 37 US cities, mail scanning, forwarding, check deposit…21 growth ideas
  • Kijimea growth strategyA German probiotics brand (founded 2011) selling K53, PRO and Women capsules; kijimea.com is its US direct-to-consumer store, run by…22 growth ideas
  • Pango growth strategyAn AI-native platform that runs e-commerce shipping, tracking and returns from one order record: delivery management, checkout delivery…22 growth ideas
  • Perchwell growth strategyA modern MLS and listings platform for real estate agents, brokerages and MLS organizations: search, listing entry, client collaboration…21 growth ideas

See every playbook and website review

Independent research. Programmable wrote this playbook from public sources, as they stood on November 10, 2026, and cites them throughout. PAX Markets did not commission, review or endorse it, and Programmable is not affiliated with PAX Markets. Company and product names are trademarks of their owners, used only to identify them. The ideas are suggestions and opinion, not professional, legal or financial advice, and not statements about how PAX Markets runs its business.

Something wrong, or do you represent PAX Markets? Email [email protected] and we will correct it or take it down.

30/60/90 roadmap

What to focus on in each month, and the ideas that carry it.

  1. Days 1 to 308 ideas

    One measured latency number, a public markets page, a fee comparison in dollars, eligibility before signup, a USDC funding guide, homepage fixes, social links and funnel tracking.

  2. Days 31 to 6011 ideas

    Liquidity and developer path: Binance migration, a λ simulator, founding market makers, the market-quality report, visible cash-back, client libraries, a trust page, introducers, routers, event qualification and an engineering blog.

  3. Days 61 to 904 ideas

    Commit to the big bets and wildcards: an independent latency audit, the equities ATS waitlist, a university λ challenge and licensing the silicon.

The research behind the plan

3 research areas, each summarised in a few lines. Open any of them for the full detail.

Customers

What customers say

Customers value speed they can measure: market makers pay for lower latency, queue priority and fast cancels. The most cited problem: crypto venue fees of 40-80 bps for small traders.

  • Zero trading fees, with cash-back instead of a commission
  • 1-10 ms round trips at crypto venues, and costly co-location
  • In their words: “tick-to-trade”

Market

Competitive landscape

4 competitors compared: Kraken Pro, Hyperliquid, EDX Markets and Binance and OKX.

  • Gap to exploit: A venue that pays ordinary traders instead of charging them
  • Kraken Pro: First-tier spot fees of 0.40% maker and 0.80% taker
  • Hyperliquid: Still charges every trader; no rebate for ordinary takers

Hiring signals

PAX is hiring only engineers, three founding roles in Redwood City on its YC job board, and the posts describe the roadmap: a Rust and SystemVerilog matching engine on AWS F2 FPGAs moving to an ASIC, BTC and ETH spot…

  • Founding Principal Hardware Engineer: Leads the in-silicon matching engine and the move from FPGA to ASIC. Ideas 1, 20, 23.
  • Founding Software Engineer: Rust hardware/software co-design; new grads welcome. Ideas 10, 14, 19, 22.
  • Founding Hardware Engineer: More FPGA/ASIC capacity; the roadmap runs from crypto to an equities ATS. Ideas 10, 21.

From the team that wrote this playbook

All products
  • ProgrammableAgencyWorking AI systems in days, not quarters (opens programmable.agency in a new tab)
  • ProgrammableInfoAn agent that finds, researches and asks first (opens programmable.info in a new tab)
  • ProgrammableDigitalScreenshots in, an animated product demo out (opens programmable.digital in a new tab)
  • ProgrammableWatchKnow the day a supplier falls out of compliance (opens programmable.watch in a new tab)
  • ProgrammableTalkGet registrants for your events, and know who’s coming (opens programmable.talk in a new tab)
  • ProgrammablePageText what you need, your customer gets a page that works (opens programmable.page in a new tab)
  • ProgrammableCommunityForums with labelled AI members, never empty (opens programmable.community in a new tab)
ProgrammableMarketing
ExamplesLeaderboardsProductsTermsPrivacyCookiesDPALegalContact

© 2026 AI-Powered Growth Intelligence