New startup ideas · Fintech · Fintech
startup idea
Marginframe
A colocated appliance that nets margin across prediction, perp and equity venues in microseconds.
Marginframe sells a 1U FPGA appliance plus support contract to brokers and venues that carry positions on more than one execution venue.
- Hardware
- Enterprise
- $1-10B market
- Creates a new category
- US first
3/5
venture judge
1
similar startups, last 2 years (1 all-time)
96%
of 4 nearest real companies still alive
sector
2 public grants in this sector, none matching the idea's terms
Test it before you build it
$2,000 · 5 weeks · 15 prospects
For $2,000 and five weeks, prove that broker treasurers posting collateral on two or more venues will pay $5,000 up front for a read-only shadow study of the margin a netted requirement would have freed, before any hardware exists.
Riskiest assumption · A broker's treasurer will pay for cross-venue margin netting on the strength of single-house collateral savings alone, before any rival venue accepts the netted risk number.
1Focus group: who and where
The treasurer, head of margin, or COO at a US FCM, broker-dealer, or proprietary trading firm that posts collateral on at least two venue types today - for example a firm carrying Kalshi event-contract positions alongside CME futures or equities - and whose excess collateral is a line item they report on monthly.
where to find 15 · The CFTC's monthly financial data for FCMs page and NFA BASIC, which together list every registered firm with its capital figures (directory); FIA Expo in Chicago this fall, where FCM treasurers and margin heads walk the floor (event); STAC Summit meetings for the low-latency trading infrastructure crowd (community); plus warm intros through investors and former colleagues into brokers already routing event contracts alongside equities, such as Interactive Brokers and Robinhood (channel).
2Sell first, build later
A four-week read-only shadow-margin study sold before the appliance exists: the broker sends daily position and margin files under NDA, and gets back a written report stating exactly how many dollars of collateral a single netted cross-venue requirement would have freed each day of the prior month, starting within two weeks of signing.
the ask · $5,000 per shadow study, prepaid, credited against the first year of a $120,000-per-year appliance license
a real yes · A real yes is a signed SOW with the $5,000 invoice paid, or a signed conditional LOI for the appliance with an install quarter; a treasurer agreeing the problem is real, an offer of free data with no payment, or a request to circle back after budget season counts as nothing.
3Small experiments
The first one attacks the riskiest assumption; each ends with a number that says whether to run the next.
1. Quantify the collateral bleed
$1,200 · 14 days
Build the target list from the CFTC FCM financial data page and NFA BASIC, filtered to firms active on multiple venue types, and book 15 treasurer calls through warm intros and one Chicago trip timed to a STAC Summit or FIA Expo. On each call, ask them to put a monthly dollar figure on collateral posted above a portfolio-level requirement, walk a two-page mock freed-collateral report, and end by pitching the paid shadow study. Founder-led, no hardware mentioned until they ask.
keep going if · 8 of 15 treasurers quantify at least $2M of idle cross-venue collateral, and 6 ask for the shadow-study SOW
2. Sell the paid shadow study
$700 · 14 days
Send the SOW to every treasurer who asked: a four-week read-only study, daily position and margin files delivered over SFTP under NDA, output is one report stating the collateral a netted cross-venue requirement would have freed, $5,000 invoiced at signing. Counsel reviews the NDA and SOW templates once; the netting math runs in software on a laptop, since the FPGA matters for latency, not for a retrospective study.
keep going if · 2 of 15 firms sign the SOW and pay the $5,000 invoice
3. Appliance LOI on study results
$100 · 7 days
Present each completed or in-flight study buyer with a one-page LOI for the 1U appliance at $120,000 per year including support, conditional on the study showing at least $250,000 in freeable collateral. Count signatures, not enthusiasm.
keep going if · 2 signed conditional LOIs with target install quarters named
4Collect a deposit up front
Tesla took $1,000 refundable reservations for the Model 3 and $100 for the Cybertruck before building either: the deposit is the measurement, not the revenue.
$5,000
per prospect, refundable
how · A paid design-partner pilot invoiced up front: the treasurer signs a two-page SOW plus NDA and pays the $5,000 study fee at signing, credited against the first-year appliance license. Chosen because regulated brokers buy analyses through SOWs routinely, while a reservation link for an uninstalled hardware box would not survive their vendor process. set up: Stripe Invoicing ↗
what it reserves · One of three shadow-study slots starting within two weeks, the credited license discount, and first position in the install queue for the appliance's pilot cohort
refund · Refunded in full if the completed study shows less than $250,000 in freeable collateral.
target · 2 paid $5,000 studies from 15 treasurer calls within 35 days
before taking money · The buyers are CFTC- and SEC-regulated brokers, so the study must stay read-only and off the order path, with their compliance team approving the NDA and data transfer before any position file moves.
Go: build it if
8 of 15 treasurers quantify $2M+ of idle collateral, 2 pay the $5,000 study fee, and 2 sign conditional appliance LOIs at $120,000 per year.
Kill: stop if
Zero paid studies after 15 treasurer calls, or fewer than 4 of 15 can put any dollar figure on cross-venue over-posting, meaning the pain is not owned or not measurable.
5 Scripts to run itoutreach message, landing copy, deposit terms · click to open
outreach message
You post full margin at every venue you clear on, and the excess sits idle while event contracts, futures and equities each demand their own collateral. I compute what one netted requirement across your venues would have freed last month - read-only, from daily position files under NDA, nothing touches your order path. The output is a single number: dollars of collateral you over-posted, day by day. Do you have 20 minutes this week to see the sample report?
landing page
How much collateral did you over-post last month? $5,000 prepaid shadow study: four weeks of daily position files in, a freed-collateral report out, credited against a $120,000-per-year appliance license Book a scoping call - three study slots before year end
deposit terms
The $5,000 study fee is invoiced when you sign the SOW and reserves one of three shadow-study slots starting within two weeks. It is credited in full against the first year of an appliance license if you proceed. If the completed study shows less than $250,000 in freeable collateral, we refund the full fee.
Would you run this test?
One tap. The yes-share feeds the Demand pillar of this idea's score; nobody sees who answered.
Budgets are out-of-pocket estimates for a team of one to three, US market. Size the deposit to the deal, and check the terms before taking money in a regulated line.
Scorecard
One score that balances how trendy the idea is, the demand for it and its potential for 100x, with competition measured relative to every other idea in the catalog. Recent startup trends first, government priorities second.
71
Idea Score, 0-100 · raw 43.9 x 1.61
Open
competition: more crowded than 6% of ideas · headwind x0.97
+0.0
government priorities, secondary (0 matching grants)
Trend
53
Is the wave forming now? 2025-26 entrants vs 2023-24, rounds since 2025, the sector's live-batch direction, the 2026 trend analyst.
- Entrants 2025-26 vs 2023-24 (similar companies)23
- Rounds announced 2025+ in the sector87
- Sector direction (live batch)50
- 2026 trend analyst50
Demand
27
Does anyone want it? YC's current RFS, companies already paid for something similar, the operator judge, founders' yes-rate in decks, readers who would run the test.
- YC asks for it (current RFS: idea / sector)30
- Someone already pays (similar companies, recent / all-time)0
- Operator judge: real pain50
100x potential
67
Can it return a fund? The venture judge (double weight), market-size and moat axes, neighbours still alive, the technologist judge.
- Venture judge50
- Market size axis33
- Moat axis100
- Technologist judge100
Score = 100 x cbrt(Trend x Demand x 100x) x (1 - 0.5 x crowding) + government bonus (max 5), calibrated so the 95th-percentile idea scores 90 (order never changes). A geometric mean: a weak pillar cannot be papered over. Percentiles are among the 272 ideas in the catalog; the terms matched were colocated, appliance, nets, margin, prediction, perp, equity, sells.
The idea in full
- What
- Marginframe sells a 1U FPGA appliance plus support contract to brokers and venues that carry positions on more than one execution venue. It ingests every venue's order and position feed, computes a portfolio risk number and an optimal collateral allocation in hardware on a deterministic clock, and returns a single netted margin requirement instead of the sum of siloed requirements. The math is an optimal-transport formulation of collateral movement, compiled to gates so the answer arrives before the next order does.
- Why now
- The radar shows four companies pitching prediction markets in two years (River Markets on prime brokerage, Oddpool on institutional infrastructure, Totalis on derivatives, Arbital on a cross-asset terminal), which means collateral is now fragmented across venue types with no shared risk layer; NSF is separately funding the underlying method with a roughly $200k award on Market Microstructure and Regularized Optimal Transport.
- Wedge: first customer and entry point
- One broker that is visibly bleeding collateral across two venues, sold founder to founder as a read-only shadow box that reports how much margin it would have freed last week before anyone routes through it.
- Path to 100x
- Trading risk and connectivity infrastructure is a $1-10B hardware and licence market, and the appliance creates a category that does not exist today: a vendor-neutral cross-venue margin box. Value compounds with each venue and broker on the frame, because every added participant increases the netting set and makes the alternative, posting full collateral at every venue, more expensive, which pushes the market to one frame.
- Ceiling
- A single large clearing incumbent or a coalition of venues builds the same netting in software and gives it away to keep flow, capping Marginframe at a niche hardware vendor with a few dozen boxes.
- Closest real companies, as the generator saw them
- The four prediction-market entrants (River Markets, Oddpool, Totalis, Arbital) are software venues, brokerages and terminals competing for traders; Marginframe sells to all of them and competes for none of their users. This is the physical risk plumbing underneath their business, and no hardware player appears on the radar.
- Main risk
- Venues refuse to accept another venue's netted risk number, so the appliance saves collateral only inside one house and never becomes a network.
Five judges
Each judge scores every idea in the catalog with a named rubric; the venture judge decides whether a card is shown at all (4-5 is venture-grade).
Venture investor
3/5
FPGA netting box with a real network effect on the collateral set, but a venue coalition giving away software netting caps this $1-10B market fast.
Bootstrapper
2/5
FPGA appliance development plus colocation for a handful of brokers is real capex before the first support contract signs.
Operator
3/5
Fragmented collateral is a measurable cost a broker's treasurer owns, but netting only pays once rival venues accept another house's risk number.
Technologist
5/5
Portfolio margin as an optimal-transport formulation compiled to FPGA gates on a deterministic clock cannot be reproduced by any prediction-market software team.
Risk
2/5
Netting only counts if rival venues accept another venue's risk number, and prediction-market venues are a regulatorily volatile buyer base.
trends
3/5
Collateral fragmentation across four new prediction-market venues (term at 6 vs 2) is a genuine 2025-2026 shift, though venues accepting a shared risk number is unproven.
Similar startups in the directory
Companies whose pitch matches most of the idea's terms (colocated, appliance, nets, margin, prediction, perp, equity, sells): 1 all-time, 1 from the last two years. Same matching as Idea Check.
We are building the financial infrastructure to let exchanges offer Perpetual Futures ("Perps") on any underlying asset.
The generator's reference companies
Real companies the model named as closest when it wrote the card, with their fate. A check mark is a company the radar could verify in its directory.
Public money in this direction
US federal grants, SBIR/STTR awards and open opportunities from the radar's public-money feed, matched to the idea's terms; the sector totals give the context.
0
grants and programs matching the idea
2
startup-relevant grants in Fintech
$809K
awarded in the sector, tracked
Market signal
What the radar sees in Fintech: new companies by cohort year, the forming YC batch, and outcomes since the February snapshot.
Fintech · 256 → 148 → 186 → 125 → 99 new companies 2022 → 2026 · 89% aliveYC F26 live: 7 in this cluster, 6% of the batch (was 5% in S26)Since February, of 592 YC companies here: 11 acquired, 8 shut down, 56 rewrote their pitch
Design attributes
The card is one cell of a designed set: every axis below was chosen before the text was written, and the text had to realize it.
- Buyer
- Enterprise
- Business model
- Hardware
- Path to 100x
- Creates a new category
- Market size
- $1-10B market
- Capital intensity
- Capital-medium (ops, field teams)
- Speed to revenue
- Revenue in 1-3 years
- Technical depth
- Deep tech: ML, hardware, bio
- Go-to-market
- Founder-led sales
- Moat
- Network effects
- Geography
- US first
- Regulation
- Unregulated
- Vibe
- Boring business
Listed under
An idea sits in its own sector and in any sector its text clearly touches.
More ideas like this
Fintech · Fintech
Bookzero
Software that closes a small company's books for a tenth of a bookkeeper's price.
Bookzero is self-serve subscription software that connects to a small business's bank feeds, invoices and payroll and produces a finished monthly close - categorized ledger, reconciliations, financial statements - for roughly $100 a month against the $800-1,500 SMBs pay outsourced bookkeepers.
Fintech · Fintech
Parbook
The self-serve secondary market where institutions trade private credit loan participations.
Parbook is a marketplace where private credit funds, insurers and regional banks list, price and settle loan participations and portfolio slices, using a standardized machine-readable data tape that Parbook generates from each seller's loan documents.
Fintech · Fintech
Perilex
An exchange where machine-priced corporate risk is listed and capital bids.
Perilex is a listing venue for idiosyncratic corporate risk that traditional carriers price badly: fleet and telematics-driven liability, compute and outage exposure, model-failure liability.
Fintech · Fintech
Countersign
Spend controls and audit trail for every dollar an enterprise's AI agents move.
Countersign sits between a company's agents and whatever rails they spend on - issued cards, wallets, stablecoin accounts - enforcing per-agent budgets, merchant allowlists and human approval thresholds before authorization, then reconciling each settled charge back to the agent run and prompt that caused it.
Fintech · Fintech
Underwatt
Underwriting software that prices AI data center risk from live power and thermal telemetry.
Underwatt sells a subscription platform to carriers, MGAs and reinsurers that prices property, business-interruption and outage risk on GPU data centers using a telemetry fleet Underwatt installs itself: power quality recorders, thermal and coolant sensors, and grid interconnect feeds at the insured sites.
Fintech · Fintech
Subrix
AI agents that settle claims between insurance carriers on one shared network.
Subrix runs autonomous agents that handle subrogation - the process where one insurer recovers claim costs from another at-fault party's insurer.
Fictional company written 2026-08-22 from MarkosWeb data; the companies, grants and numbers around it are real and tracked. Treat the idea as a research prompt, not a plan.