MarkosWebWeb intelligence for the agentic economy

New startup ideas · Gifted kids, funded by proof · Integrity: fake results, fake need, and child safety

startup concept

Fairform

Catches fabricated records and AI-written applications before an SGO funds them

Application-integrity software for scholarship-granting organizations: it cross-references claimed achievements against registries, flags coach recommendation letters reused across applicants, and detects AI-ghostwritten essays and coached applications in the pool.

2

similar startups, last 2 years (3 all-time)

yes

8 matching federal grants and programs

Direction supported by government programs and grants

Test it before you build it

$600 · 4 weeks · 15 prospects

Proves for $600 in 4 weeks that SGO program directors will pay $3,000 for fraud screening of a real applicant pool before the 2027 surge, despite the false-positive risk.

Riskiest assumption · A program director at a mid-sized SGO believes fabricated applications will contaminate their 2027 pool badly enough to pay for screening now, even knowing that one false flag against a genuine low-income kid could publicly embarrass the organization.

1Focus group: who and where

The program director or director of operations at a mid-sized scholarship-granting organization (500-10,000 applications a year, 5-30 staff) in an opted-in state, currently planning a 2027 intake they expect to be several times 2026 volume with no way to verify claimed achievements.

where to find 15 · Certified SGO lists published by state agencies (Arizona Department of Revenue's certified School Tuition Organization list, Pennsylvania DCED's EITC scholarship organization list, Florida DOE's list of eligible SGOs); the National Scholarship Providers Association member directory and its October 2026 annual conference; EdChoice's directory of school choice programs to map which of the 27 opted-in states have mid-sized SGOs worth calling.

2Sell first, build later

A retrospective integrity audit of one past application cycle (up to 300 applications): per-application evidence reports covering duplicate and reused recommendation letters, registry mismatches on claimed olympiad and academic results, and essay-authorship anomalies, delivered within 21 days of file handoff, before the SGO's 2027 intake opens.

the ask · $3,000 per audit, $1,000 invoiced up front as a deposit, balance on delivery

a real yes · A real yes is a signed one-page audit agreement plus the $1,000 deposit paid; a scope call, a compliment on the sample report, or an offer to pilot for free counts as a no

3Small experiments

The first one attacks the riskiest assumption; each ends with a number that says whether to run the next.

  1. 1. Pitch paid retrospective audits

    $250 · 14 days

    Build a list of 40 SGOs from the state lists and NSPA directory, email the program director, and book 15 twenty-minute calls. Ask how they verify claimed achievements and letters today and whether they have caught a fabricated application; end every call by offering a $3,000 retrospective integrity audit of a past cycle's pool, delivered as per-application evidence reports in 21 days. One founder runs all calls with a one-page sample report as the artifact.

    keep going if · 8 of 15 directors name fabricated applications or AI-written essays as a top-3 worry for 2027, and 3 of 15 agree to receive a written audit scope

  2. 2. Borderline-case report walkthrough

    $150 · 7 days

    Hand-build a mock evidence report for a fictional pool of 10 applicants that includes one deliberately borderline flag (a genuine low-income kid whose registry record is incomplete). Walk the 5 most interested directors through it and watch whether the borderline case makes them recoil or ask for the workflow. This tests whether the false-positive fear kills the purchase before it starts.

    keep going if · 4 of 5 directors still want the audit after seeing the borderline case, and can describe how their reviewers would handle an amber flag

  3. 3. Collect audit deposits

    $200 · 10 days

    Send the directors who agreed to a scope a one-page e-signed agreement and an invoice for a $1,000 deposit against the $3,000 audit, with a named delivery date 21 days after file handoff. No software is built; the first audits run on scripts and manual registry lookups by the founders.

    keep going if · 2 of the scoped SGOs pay the $1,000 deposit within 10 days of receiving the invoice

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.

$1,000

per prospect, refundable

how · A $1,000 deposit invoiced up front on the $3,000 audit under a one-page e-signed design-partner agreement, signed by the SGO's executive or program director and paid by check or ACH - an invoice matches how a nonprofit actually moves money, so a card checkout would only add friction. set up: Stripe Invoicing ↗

what it reserves · One of five February 2027 audit slots, a delivery date 21 days from file handoff, and first-cohort pricing on the 2027 intake screening product

refund · Refunded in full if the evidence reports are not delivered within 21 days of receiving the files.

target · 2 paid deposits from 15 conversations within 28 days

before taking money · Application files contain minors' records, so run every audit under a signed data-processing agreement on files the SGO has pseudonymized, and never contact an applicant or label a named child fraudulent.

Go: build it if

2 SGOs pay the $1,000 deposit within 4 weeks, at least 8 of 15 directors name fabrication as a top-3 problem for 2027, and 4 of 5 accept the mock report after seeing the deliberate borderline case.

Kill: stop if

0 deposits after 15 completed conversations, or fewer than 5 of 15 directors rank fabrication in their top three problems, or 3 of 5 walk away from the report over the borderline case - that last one confirms the stated risk is fatal.

5 Scripts to run itoutreach message, landing copy, deposit terms · click to open

outreach message

You certified as an SGO knowing your 2027 intake will be several times this year's pool, and by January the AI-written essays and recycled coach letters in it will read exactly like the real ones. I run retrospective integrity audits: give me one past cycle's applications and in 21 days you get a per-application evidence report - duplicate letters, registry mismatches, essays that do not match the applicant. $3,000, five slots before intake opens. Do you have 20 minutes this week to look at a sample report?

landing page

Know which scholarship applications are fabricated before you fund them $3,000 audits one past cycle: duplicate letters, registry mismatches, per-application evidence reports in 21 days Reserve one of five February audit slots with a $1,000 deposit

deposit terms

Your $1,000 deposit reserves one of five February 2027 audit slots and is credited against the $3,000 audit fee. It is refunded in full if your evidence reports are not delivered within 21 days of file handoff. The balance is invoiced on delivery, before your 2027 intake opens.

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

Ranked against every idea in the catalog: trend, demand and 100x potential from the corpus, competition relative to the other ideas. A generated concept has no judges or swipes yet, so its pillars use the data signals only.

74

Idea Score, 0-100 · raw 45.8 x 1.61

Warm

competition: more crowded than 44% of ideas · headwind x0.78

+3.1

government priorities, secondary (13 matching grants)

Trend

45

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)92
  • Rounds announced 2025+ in the sector44
  • Sector direction (live batch)0

Demand

47

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)60
  • Someone already pays (similar companies, recent / all-time)33

100x potential

78

Can it return a fund? The venture judge (double weight), market-size and moat axes, neighbours still alive, the technologist judge.

  • Neighbours still alive78

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 382 ideas in the catalog; the terms matched were catches, fabricated, records, ai-written, applications, sgo, funds, application-integrity.

The concept in full

What
Application-integrity software for scholarship-granting organizations: it cross-references claimed achievements against registries, flags coach recommendation letters reused across applicants, and detects AI-ghostwritten essays and coached applications in the pool. Reviewers get a per-application evidence report instead of a gut call. Because it scores evidence rather than polish, it protects exactly the kids who cannot afford application consultants.
Grounded in (2025-2026 signals)
The One Big Beautiful Bill Act (July 4, 2025) created the up-to-$1,700 federal credit for gifts to SGOs; on June 8, 2026 the IRS reported 27 states had elected to participate, Texas, Florida, Ohio, Virginia, Colorado and Tennessee among them; YC's Fall 2026 RFS includes Proving You're Human, and detecting machine-written scholarship applications is that problem applied to K-12 funding.
What it rides
Federal scholarship tax credit from 2027, 27 states in, as the fraud surface: a nonrefundable credit that pays donors back attracts organized abuse, and SGOs must show objective, nondiscriminatory selection to keep their status
Why now
With 27 states in as of June 8, 2026 and the credit claimable from January 1, 2027, SGO application volume will jump in one cycle, and the same year AI essay tools make a coached application indistinguishable from a real one by reading alone; the YC Fall 2026 RFS Proving You're Human confirms detection is a live category.
Wedge: first customer and entry point
First customer is one mid-sized SGO in an opted-in state like Florida or Texas preparing its 2027 intake, starting with academic and olympiad scholarship pools where claims are checkable. A 2-person team ships duplicate-letter and registry-mismatch detection first, essay forensics second.
Closest real companies, as the generator saw them
Edexia and Frizzle (both YC 2025) apply AI to grading student work for teachers, not to fraud in funding decisions; Alice.tech (YC W25) does exam prep. No 2025-2026 company in the briefs works on scholarship application integrity.
Main risk
False positives that wrongly tag a genuine low-income kid as fraudulent would be catastrophic for an SGO, so one public mistake kills trust in the whole approach.

Similar startups in the directory

Companies whose pitch matches most of the concept's terms (catches, fabricated, records, ai-written, applications, sgo, funds, application-integrity).

  • Florinyc S26 · 2026 · Fintechalive

    Florin is an AI-native insurance company that runs autonomously

  • GeniPhysplugandplay PnP 2025 · 2025 · Healthcare and bioalive

    GeniPhys is developing and commercializing its patented Collymer™ technology platform: a first-in-class polymerizable (scaffold-forming) collagen that supports custom design and fabrication of collagen materials for accelerated and improved tissue restoration and regeneration.

  • Meticulousyc S21 · 2021 · Developer toolsalive

    Automated e2e screenshot testing without writing or maintaining tests

Public money in this direction

US federal grants and open opportunities matched to the concept's terms.

Other concepts in this collection

Fictional concept generated 2026-08-27 by claude-fable-5 from the collection's brief and MarkosWeb data. Treat it as a research prompt, not a plan.