New startup ideas · Consumer and commerce · Consumer
startup idea
Alcuin
AI tutoring agents that give every enrolled student a personal tutor at 10x lower cost.
Alcuin sells AI tutoring agents to universities, textbook publishers, and online course operators, who deploy them across whole cohorts so every student gets unlimited one-on-one tutoring at roughly a tenth of what human tutoring lines cost institutions today.
- AI agent as a service
- Enterprise
- $100B+ market
- Collapses a cost 10x
- Global from day one
3/5
venture judge
7
similar startups, last 2 years (52 all-time)
92%
of 2 nearest real companies still alive
yes
8 matching federal grants and programs
Direction supported by government programs and grants
Test it before you build it
$1,500 · 6 weeks · 20 prospects
For $1,500 in six weeks, prove three deans of online programs will sign paid spring-term tutoring pilots now instead of waiting for their LMS vendor or a foundation model.
Riskiest assumption · A dean of online learning with a documented DFW problem will commit budget this term to a startup's tutoring pilot rather than wait for their LMS vendor or a foundation model provider to ship tutoring natively.
1Focus group: who and where
The dean or associate dean of online and digital learning, or the director of student success, at a US university or community college with 5,000 or more online enrollments, staring at 25-40 percent DFW rates in college algebra, intro statistics, or anatomy while their human tutoring line cannot scale to those cohorts.
where to find 20 · The UPCEA and WCET member directories (the associations online-learning leaders actually belong to), an IPEDS-derived list of the US institutions with the largest exclusively-online enrollments, and the OLC Accelerate 2026 speaker list plus EDUCAUSE community groups for warm entry through people they already know.
2Sell first, build later
A spring 2027 term pilot: tutoring agents deployed on the program's two highest-DFW gateway courses for every enrolled student, with mastery reports for instructors and a retention delta measured against the prior three terms - published as a joint case study or kept private, the dean's choice.
the ask · $10,000 per pilot per term; $2,500 due at signing.
a real yes · A real yes is a signed pilot agreement with a dated start plus a $2,500 deposit paid or a PO issued; enthusiastic committee meetings, unpaid pilots, and send-it-to-our-AI-task-force are not.
3Small experiments
The first one attacks the riskiest assumption; each ends with a number that says whether to run the next.
1. Dean pain-and-timing calls
$300 · 14 days
Build a list of 30 institutions from IPEDS enrollment data cross-referenced with UPCEA and WCET member directories, and book 15 calls through warm intros and direct email. Map their two worst-DFW gateway courses, what the tutoring line costs per student today, and who can sign a sub-$10,000 pilot without a committee. Founder runs every call.
keep going if · 8 of 15 deans name a specific course, a current tutoring cost, and this-year budget authority, and 6 request the priced pilot scope
2. Pilot-now versus wait choice
$500 · 14 days
Show a scripted demo of the tutoring agent running on their own public syllabus, assembled with off-the-shelf models and no production code, then present two options in writing: a $10,000 spring 2027 pilot on their two worst courses, or a free notify-me list for whenever the big vendors ship this. The split directly measures the wait-for-foundation-models objection.
keep going if · 5 of 12 deans who see the choice pick the paid-now track and bring IT or procurement into the next call
3. Signed pilot and money moved
$700 · 21 days
Close with a pilot agreement plus a FERPA data processing agreement (templates with one attorney review pass), taking $2,500 at signing or an issued PO, with a dated spring 2027 kickoff. Walk each paid-track dean through signature and their IT review inside three weeks.
keep going if · 3 signed pilot agreements with dated spring starts, at least 2 with the deposit paid or a PO issued
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.
$2,500
per prospect, refundable
how · A signed pilot agreement with $2,500 invoiced at signing, priced deliberately below most institutions' competitive-bid threshold so a dean can use discretionary budget; where even that stalls, a countersigned agreement with a dated spring start and a PO under procurement review counts as the commitment, and a verbal yes does not. The dean signs the pilot, IT signs the FERPA DPA. set up: Stripe Invoicing ↗
what it reserves · One of three spring 2027 pilot slots, the named course pair, and a kickoff date aligned to the registrar's term calendar.
refund · Refunded in full if the pilot has not kicked off by the agreed start date or the institution cancels before kickoff.
target · 3 signed pilots with money moved or PO issued from 20 dean conversations within 6 weeks
before taking money · Do not let agents touch live students or real student records before the institution signs the FERPA data processing agreement and the pilot clears its IT and security review; the deposit buys a slot, not early access to student data.
Go: build it if
3 signed pilot agreements with dated spring 2027 starts, at least 2 with $2,500 paid or a PO issued, out of 20 conversations - build the pilot deployment for those named courses.
Kill: stop if
After 20 conversations, no money moved, no PO, and fewer than 3 institutions take the priced scope into IT or procurement review - buyers are waiting for foundation models exactly as the card fears; stop.
5 Scripts to run itoutreach message, landing copy, deposit terms · click to open
outreach message
Your two highest-DFW gateway courses are where your online program loses the most students and the tuition that leaves with them. I build tutoring agents trained on your own curriculum that give every enrolled student unlimited one-on-one help at about a tenth of what your tutoring line costs per student today, and I'm signing three $10,000 spring-term pilots - retention delta published or kept private, your call. Do you have 20 minutes this week to see whether your algebra or stats sequence fits?
landing page
Cut DFW rates in your gateway courses with unlimited one-on-one tutoring $10,000 spring-term pilot on two courses - $2,500 at signing reserves a slot Book a 20-minute scoping call - three pilot slots for spring 2027
deposit terms
$2,500 at signing reserves one of three spring 2027 pilot slots for your named course pair and a kickoff date on your term calendar, credited against the $10,000 pilot fee. Refunded in full if the pilot has not kicked off by the agreed start date or if your institution cancels before kickoff. The remaining $7,500 is invoiced at kickoff.
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.
62
Idea Score, 0-100 · raw 38.7 x 1.61
Active
competition: more crowded than 65% of ideas · headwind x0.67
+3.9
government priorities, secondary (45 matching grants)
Trend
38
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)3
- Rounds announced 2025+ in the sector38
- Rounds announced 2025+ matching the idea99
- Sector direction (live batch)0
- 2026 trend analyst50
Demand
70
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)100
- Operator judge: real pain50
100x potential
52
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 axis100
- Moat axis20
- Neighbours still alive65
- Technologist judge25
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 tutoring, enrolled, student, personal, tutor, sells, universities, textbook.
The idea in full
- What
- Alcuin sells AI tutoring agents to universities, textbook publishers, and online course operators, who deploy them across whole cohorts so every student gets unlimited one-on-one tutoring at roughly a tenth of what human tutoring lines cost institutions today. The agents are built on pedagogy-tuned models with mastery estimation and retrieval over the institution's own curriculum - deep ML work, but capital-light because the product is pure software. Sales are founder-led into education buyers whose procurement, accreditation, FERPA, and EU AI Act high-risk obligations make this a heavily regulated, 1-3 year revenue path from day one of a global rollout.
- Why now
- Medly AI raised $8M on 2026-08-19 to expand AI-powered tutoring for students, and YC's Fall 2026 RFS includes The Primer - the capital and the demand signal for AI-native education are both on the table, but the funded players are consumer apps, leaving the institutional deployment layer open.
- Wedge: first customer and entry point
- One founder-led deal with a single online degree program: deploy tutoring agents on its two highest-failure-rate courses and publish the retention delta, then walk that case study into publishers.
- Path to 100x
- Global education and tutoring spend is a $100B+ market, and an agent that replaces a per-hour human cost with a per-student software fee collapses that cost 10x - the classic mechanism by which a services line becomes a software category. The 100x case is becoming the default tutoring layer for institutional education worldwide, with outcome data and curriculum integrations hardening into the moat the company admittedly lacks today.
- Ceiling
- Education procurement is slow and consensus-driven, and if adoption stays pilot-sized the company stalls as a niche courseware vendor.
- Closest real companies, as the generator saw them
- Medly AI and Wondering sell tutoring and learning directly to consumers; Alcuin sells cohort-wide deployments to institutions, where the buyer is the provost or publisher, not the student. Doomersion is a consumer language-learning wedge, not an institutional agent.
- Main risk
- Foundation model providers ship credible tutoring behavior natively, and institutions wait for that instead of buying from a startup with no moat yet.
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
$100B+ education spend and capital-light margins, but the card concedes no moat yet against foundation models shipping tutoring natively.
Bootstrapper
3/5
Capital-light software with real institutional budgets, but provost-level procurement, FERPA and EU AI Act obligations stretch the first meaningful contract past a year.
Operator
3/5
Institutions do pay per-hour tutoring lines today, but consensus procurement plus waiting for foundation models blunts a buyer who feels no personal pain.
Technologist
2/5
Pedagogy prompting plus retrieval over curriculum is a wrapper the foundation model providers can absorb, and the card concedes no moat yet.
Risk
2/5
No moat by its own admission, EU AI Act high-risk obligations, and foundation model providers can ship credible tutoring natively.
trends
3/5
The Primer RFS is a real anchor, but institutional AI tutoring was already sellable in 2023 and the card concedes no moat against native model tutoring.
Similar startups in the directory
Companies whose pitch matches most of the idea's terms (tutoring, enrolled, student, personal, tutor, sells, universities, textbook): 52 all-time, 7 from the last two years. Same matching as Idea Check.
AI tutor for each student.
AI tutors for exam prep
AI tutor starting with maths and science
Helping teachers and students use AI to enhance learning
Every student's personalized learning assistant
Personalized learning built for schools
An AI tutor that helps you study
Studeo Limited is democratising elite tutoring.
Online University for Tech Professionals
Reading + Writing Club App for Kids
We help students in Latin America graduate from College.
Kunduz provides instant answers to students' questions and makes…
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.
- Medly AI
- Wondering ✓ 2026
- Doomersion ✓ 2026
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.
8
grants and programs matching the idea
33
startup-relevant grants in Consumer
$10M
awarded in the sector, tracked
1
opportunities open now in the sector
- SBIR Phase II: Colleague AI: Building Next-Generation K-12 STEM Learning Environments through Specialized AI Modelsawardhigh relevance
National Science Foundation · SBIR Phase II · $1M · posted 2026-07-28
- Auxigen: BIO – An Adaptive, AI-Driven Commercialization Training Platform for Biomedical Researchersawardhigh relevance
NIH / NIGMS · SBIR phase I · $309K · posted 2026-09-01
- I-Corps: Translation potential of a virtual coach-guided wellness platform designed to improve college student well-being, academic success, and retentionawardhigh relevance
National Science Foundation · I-Corps · $50K · posted 2026-08-19
- VINES: Track 1: NSF-JST: SENRIGAN: Infrastructure-based autonomous driving for micromobilityawardhigh relevance
National Science Foundation · Use-Inspired NextG, GVF - Global Venture Fund · $675K · posted 2026-07-30
NIH / NEI · SBIR phase I · $307K · posted 2026-07-15
- Context-Aware Conversational Tutors for Longitudinal Learning in Engineering Educationawardmedium relevance
National Science Foundation · IUSE · $750K · posted 2026-07-29
National Science Foundation · IUSE · $750K · posted 2026-08-03
- CyberAI Innovation: Training the Human Firewall Through Generative Artificial Intelligence Adversaries and Metacognitive Scaffolding for Cybersecurity Workforce Resilienceawardmedium relevance
National Science Foundation · CyberAICorps SFS · $500K · posted 2026-08-01
Market signal
What the radar sees in Consumer: new companies by cohort year, the forming YC batch, and outcomes since the February snapshot.
Consumer · 182 → 222 → 211 → 156 → 64 new companies 2022 → 2026 · 86% aliveYC F26 live: 4 in this cluster, 3% of the batch (was 3% in S26)Since February, of 782 YC companies here: 6 acquired, 4 shut down, 45 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
- AI agent as a service
- Path to 100x
- Collapses a cost 10x
- Market size
- $100B+ market
- Capital intensity
- Capital-light (software margins)
- Speed to revenue
- Revenue in 1-3 years
- Technical depth
- Deep tech: ML, hardware, bio
- Go-to-market
- Founder-led sales
- Moat
- No moat yet
- Geography
- Global from day one
- Regulation
- Heavily regulated
- Vibe
- Hot space
Listed under
An idea sits in its own sector and in any sector its text clearly touches.
More ideas like this
Consumer and commerce · Consumer
Yearmark
Certified on-device age assurance and minor-safety controls, shipped as one SDK for consumer apps.
Yearmark is a subscription SDK that consumer app companies drop into signup and payment flows to estimate a user's age band on-device from face, voice and interaction signals, then enforce the right minor-safety rules per jurisdiction without sending biometrics to a server.
Consumer and commerce · Consumer
Talvia
FDA-cleared autonomous AI therapy, sold to health plans at one-tenth the cost.
Talvia builds a clinically validated conversational AI that delivers protocolized CBT for anxiety and depression as a covered insurance benefit, replacing 150-dollar-per-session human therapy with a 15-dollar-per-member-per-month product.
Consumer and commerce · Consumer
Fidara
A licensed AI fiduciary that manages your entire financial life, legally on your side.
Fidara is a consumer AI agent registered as an investment advisor and licensed money transmitter, so it can actually move money, rebalance accounts, refinance debt, contest bills and file the paperwork - not just suggest.
Consumer and commerce · Consumer
Perpetua
A marketplace for persistent AI worlds that remember every player who visits.
Perpetua trains a real-time interactive world model with durable state, so a generated world keeps its geography, characters and consequences between sessions and across different visitors.
Consumer and commerce · Consumer
Castmark
Global marketplace where brands license AI-ready characters, voices and likenesses per use.
Castmark lists characters, voice models and performer likenesses with machine-readable licence terms (territory, media, duration, prohibited uses) and sells them per generation or per campaign to game studios, publishers and consumer brands.
Consumer and commerce · Consumer
Vaultline
Instant marketplace for vaulted collectibles: buy, sell and settle without ever shipping.
Consumers deposit graded cards and sealed hobby product into a partner bonded vault once; after that, ownership transfers instantly on Vaultline's ledger and nothing moves physically until an owner requests withdrawal.
Fictional company written 2026-08-26 from MarkosWeb data; the companies, grants and numbers around it are real and tracked. Treat the idea as a research prompt, not a plan.