Bloomy
AI tutoring app that gates K-12 students through mastery-based skill graphs.
The call
Bloomy is a small, fast-moving wrapper on shared AI models racing a free-to-cheap nonprofit rival with 1,400x its reach.
Bloomy builds an AI tutor for K-12 students using the same AI models (from OpenAI and Anthropic) that anyone else can license. Khan Academy already launched Khanmigo, a nearly identical tutor that costs 20 times less per student, reaches 1.4 million users across 380 school districts, and has no reason to lose students to a smaller rival since Khan is a nonprofit. Bloomy's tutoring software is a thin layer on top of shared AI technology, with no special student data, locked-in workflows, or legal protection that would make it hard to copy or replace. The company's founder has strong credentials in education and strategy, but is running solo while better-funded competitors like Synthesis already have millions in revenue and statewide adoption. What could shift this: if Bloomy proves in a real controlled study that students using it actually score higher on standardized tests, lands a major school district contract, or builds a team fast enough to scale before Khan or the AI companies themselves add more tutoring features.
Why
01
AI pressure: OpenAI and Anthropic native tutoring/study features. Bloomy's tutor runs on borrowed OpenAI and Anthropic models, so those same labs can ship the tutoring layer directly and skip Bloomy. Could AI do the core job? Yes, likely.
02
Bigger players: Khan Academy's Khanmigo. Khanmigo already serves 700k+ students across 380+ districts for about $15 a year, versus Bloomy's ~500 students at $279 a year. Could a giant copy it? Yes, easily.
03
What protects it. A mastery-gated skill graph built with Learning Commons/CZI input and early pilot NWEA gains that a generic tutor bot doesn't replicate yet.
What would change this call
Bloomy reports actual Fall 2026 enrollment numbers versus its 'thousands of new students' claim. Shows whether the product is gaining real traction beyond its initial ~500-student pilot.
Publication of a controlled, third-party NWEA study on Bloomy's learning gains. Would show whether the 1.8x pilot result is causal and repeatable, not a small-sample fluke.
Khan Academy or Synthesis announces new pricing, features, or district deals in K-12 AI tutoring. Shows whether larger rivals are closing the gap or squeezing Bloomy's window further.
Publication of a controlled, third-party NWEA study on Bloomy's learning gains. Would show whether the 1.8x pilot result is causal and repeatable, not a small-sample fluke.
Khan Academy or Synthesis announces new pricing, features, or district deals in K-12 AI tutoring. Shows whether larger rivals are closing the gap or squeezing Bloomy's window further.
Key numbers
- ~500 students across 6 schools
- Pilot showed ~1.8x expected NWEA growth, 150 students
- ELA priced at $279 per learner per year
Named risks
- Core BloomyBot tutor is a thin wrapper on shared Anthropic/OpenAI models with no exclusive data or lock-in.
- Khan Academy's free/near-free nonprofit Khanmigo offers identical mastery-tutor mechanics at far larger scale.
- Solo-founder team with only one hire in progress, limiting execution capacity against funded rivals.
- Revenue and growth claims (ARR doubling every 2 weeks) are entirely self-reported and unverified.
- NWEA growth pilot is non-randomized, single-school, small sample, not yet a controlled study.
Rivals, as read in July 2026
| Rival | Direction | Evidence |
|---|---|---|
| Khan Academy (Khanmigo) | losing to | Khanmigo hit 700k+ students, 380+ districts, $15/yr pricing vs Bloomy's ~500 students at $279/yr, mid-2026 |
| Synthesis Tutor | losing to | Synthesis on pace for $10M+ 2025 revenue, $17.5M raised, statewide Oklahoma adoption 2025-26 |
| Carnegie Learning (MATHia) | contesting | MATHia cites 12-15% proficiency gains in independent studies, per June 2026 report |
| MagicSchool AI | losing to | MagicSchool reached 6 million educator users by October 2025, though targets teachers not students |
Questions people ask
- Is Bloomy profitable in 2026?
- No public profitability data exists. Bloomy is a YC S26 company founded by Alex Southmayd, built between February and April 2026, with founder-reported ARR that 'roughly doubled every 2 weeks' since starting YC in June 2026, but no absolute revenue or profit figures have been disclosed as of July 2026.
- Could AI replace Bloomy?
- Bloomy's entire product, diagnosis, lesson pathing, and Socratic tutoring, is a scaffolded wrapper on shared Anthropic and OpenAI models as of July 2026. With no exclusive data layer, it is exposed to frontier labs or Khan Academy shipping equivalent native tutor features directly.
- Who is Bloomy's biggest competitor?
- Khan Academy's Khanmigo, which reached 700,000+ K-12 students and 380+ district partners in 2024-25, growing to 1.4 million users by mid-2026, priced around $15/student/year for districts versus Bloomy's $279/year ELA product.
- Who owns Bloomy?
- Bloomy is privately held, founded and solely run by Alex Southmayd (ex-Teach For America, McKinsey, Stanford GSB MBA), and is part of Y Combinator's Summer 2026 (S26) batch, with no disclosed outside seed round beyond standard YC investment as of July 2026.
- How many students does Bloomy have?
- Bloomy reached approximately 500 students across 6 schools between April and June 2026, with the founder expecting 'thousands' of new students by fall 2026, per a July 21, 2026 job listing; this remains self-reported and unverified.
- Does Bloomy have proof its product improves learning outcomes?
- A non-randomized pilot at a Massachusetts charter school (~150 students, grades 6-8) showed students averaging roughly 1.8x expected winter-to-spring NWEA MAP growth, disclosed July 20, 2026; the founder frames this as observational, not causal proof.
Timeline
- 2026-07-31Bloomy scheduled to launch its Math product at ELA's price point
- 2026-07-21Bloomy posts founding engineer role at $175K-$200K plus 1.5-2.0% equity
- 2026-07-20Bloomy publicly launches on Hacker News, discloses pricing and pilot NWEA data
- 2026-06Alex Southmayd enters Y Combinator's Summer 2026 (S26) batch
- 2026-04Bloomy reaches roughly 500 students across 6 schools by this point (through June)
- 2026-02Southmayd begins building Bloomy's product, through April 2026
Revision history
27 July 2026
First assessed. Losing its protection. The text above is fixed as of this date and is not edited afterwards.
In the same band
Sources
- The EdTech AI Visibility Index 2026 (2026-07-07)
- Best AI Tutoring Software for School (2026-06-08)
- Best AI Education Tools in 2026 (2026-06-15)
- (job listing feed) (2026-07-21)
- Launch HN: Bloomy (YC S26) (2026-07-20)
- 3D Math and ELA Game for K-6 Kids
- Bloomy: AI-powered mastery learning for K-12
- Introducing study mode (2025-07-29)
- What makes Khan Academy a not-for-profit?, Khan Academy Help Center (2021-09-20)
- Khan Academy - Funding