Building the intelligence layer for the next generation — across every subject, personalized to every mind.
Pragnya is not EdTech in the content-delivery sense. It is an intelligence cultivation platform — building the reasoning, transfer, and retention capacity of students across subjects through AI-driven Socratic sessions, personalized learning paths, and outcome-verified learning cycles.
The post-BYJU's generation of schools and parents has one demand: prove it worked. We are building the only platform that measures, improves, and proves how deeply students think — not just what they completed.
The problem is not content. It is the absence of thinking.
Why content-heavy EdTech fails
Billions have been invested in video libraries, drill platforms, and generic AI tutors. Students have more content than ever — and reasoning outcomes have not improved. Content is not intelligence. The market learned this from BYJU's. The next wave must build differently.
Why personalization has not worked yet
Adaptive platforms personalize content sequence. No platform personalizes the thinking challenge — the exact Socratic question that pushes this specific student's reasoning one level deeper, based on their actual thought process. That is what Pragnya does.
Why now
LLMs make it possible — for the first time — to score open-ended reasoning at scale, generate personalized Socratic probes, and identify specific misconceptions in real time. This was not economically viable 3 years ago. The timing is right, and the India market is waiting for an outcome-proven platform.
A category that does not yet have a clear winner
India K-12 — initial wedge
Post-BYJU's, progressive schools and parents are actively seeking outcome-proven alternatives. India's K-12 EdTech market is projected at $30B by 2030. The skepticism of the BYJU's era creates a unique trust advantage for a platform that leads with evidence.
Direct learner market
Parents of high-potential underperformers are willing to pay for personalized thinking development outside the school system. Pragnya's multi-subject, interest-driven model is a natural fit for this segment — and does not require school procurement cycles.
Global AI-in-Education
AI in Education is growing at ~40% CAGR, projected to reach $25B by 2030 globally. Pragnya's architecture — Go LI agent + Claude API + FSRS scheduler — is built to scale multi-language, multi-curriculum from the start.
What no funded player currently owns
Khan Academy tutors content. Century Tech adapts content sequence. Carnegie Learning teaches math. Nobody owns the reasoning quality layer — the measurement and cultivation of how students think. That is Pragnya's category.
Step-by-step reasoning record from every session — a proprietary data asset no competitor can buy or replicate.
Gold-standard cognitive science metric — commercially absent. Every session ends with a transfer challenge.
Cross-subject reasoning compounds. Math logic sharpens argument analysis. No competitor spans all subjects.
We personalize the thinking challenge — not the content sequence. A fundamentally different layer.
30/60/90-day retention checks. Transfer scores per concept. Proof that thinking quality improved — not hours logged.
What we measure — direct, defensible, science-grounded
Three revenue streams, one compound advantage
Pilot → Subscription
Paid diagnostic pilots establish baselines. Schools that see quantified LI improvement convert to annual subscriptions for ongoing measurement, cohort reports, and mentor-flagging dashboards. Pilot-first reduces CAC and builds trust before any recurring commitment.
Direct Subscription
Monthly/annual subscription for students and parents outside the school system. Includes personalized reasoning sessions, multi-subject LI profile, retention checks, and mentor access tiers. Natural retention: intelligence growth is visible, measurable, and addictive.
API & Insights (Year 2+)
Thinking trace data — anonymized and aggregated — is a compounding research asset. Long-term: LI-as-a-Service API for other education providers, curriculum research institutions, and assessment organizations who lack the infrastructure to measure reasoning quality at scale.
Where we are today
Infrastructure built
Identity service (Go/Gin), community/cohort service (Python/FastAPI), Flutter app shell, Keycloak SSO integration — all production-grade with structured logging, typed errors, and test coverage. The foundation is ready; the LI engine is the critical next build.
Pilot-ready framework
4-step LI pilot playbook operational: baseline diagnostic → mentor-led cohort → post-diagnostic → retention checks. Early mentor cohorts running with feedback loops actively refining intervention protocols.
Dashboard intelligence live
LLM-powered dashboard recommendation engine deployed — generates today's learning action with explainability for students, parents, and mentors. Awaiting real session data from the LI engine to move from deterministic fallback to live personalized recommendations.
From pilot to platform to intelligence-at-scale
LI Engine + Reasoning Canvas
- pragnya-li-engine (Go + Claude API)
- Socratic session + step scoring
- FSRS spaced repetition scheduler
- Flutter Reasoning Canvas screen
- Fix form → Formspree endpoint
Multi-Subject + Personalization
- 6-subject problem bank (Maths, Science, Language, Social, Economics, Biology)
- Per-subject intelligence score + Bloom's tracker
- Interest-based problem context matching
- Transfer challenge generation (Claude API)
- Parent and mentor dashboards
School Pilots + Revenue
- Self-service pilot onboarding for schools
- Pre/post LI outcome report (PDF + web)
- Cohort misconception flagging dashboard
- First paid subscriptions (schools + direct learners)
- Mobile app (Flutter iOS + Android)
Scale + Series A
- Expand beyond India: SE Asia, Middle East
- LI-as-a-Service API for third-party platforms
- Cross-subject intelligence graph (concepts connecting across domains)
- Series A with outcome data from 10,000+ learner sessions
Built by people who care about how minds grow
Leads LI platform strategy, school pilot design, outcome framework, and engineering architecture. Built the identity, community, and auth infrastructure. Driving the LI engine and reasoning canvas development.
Leads product delivery, pilot execution, school reporting workflows, and partnership development. Manages the mentor cohort operations and outcome verification process.
Lead cohort delivery, mentor training, and evidence-backed learning intervention design. Ensure the human layer of the platform is as precise as the AI layer.
Full team and advisor bios: team.html
Pre-seed round — building the intelligence engine
We are raising pre-seed capital to build the pragnya-li-engine (Go + Claude API Socratic agent), productize multi-subject reasoning sessions, expand mentor-led school pilots, and scale go-to-market to schools and direct learners.
This is a pre-seed bet on the category: personalized intelligence cultivation, not content delivery. The infrastructure is built. The moat is the thinking trace dataset that grows with every session. We are looking for investors who understand why reasoning quality is the next frontier after content access.
Detailed financials, pilot data, and cap table available on request under NDA.