A production-grade platform, with a lot of shipped work still to unlock
WeLearnYouLearn already runs five portals on one shared data model for Indian schools. Here’s the technical case, honestly, including what’s still ahead of us.
Portals sharing one data model
Database migrations shipped
Database indexes, incl. tenant-scoped
AI functions built (1 live today)
Market
India's schools are still running on spreadsheets and paper
Attendance registers, fee ledgers and syllabus tracking are frequently manual or fragmented across disconnected tools — even where basic software exists, it rarely spans admin, teacher, student and parent in one system.
Technical differentiation
What’s actually underneath the product
Every request checks school_id before touching data — isolation enforced in one place, not scattered across routes.
Multi-tenant from the ground up
Every school's data is scoped by school_id and checked in application code on every request, centralized in one guard function rather than scattered ad hoc. Cross-school access — accidental or malicious — is explicitly rejected.
A schema that's actually been through production cycles
~221 idempotent migrations and 99 indexes, including tenant-scoped composite indexes and full-text search — the signature of a schema that's evolved under real feature pressure, not a weekend prototype.
Security treated as an ongoing practice
Separate, isolated login sessions per portal. Bcrypt password hashing. Two real commits in project history are literally titled 'add missing auth/tenant checks' — evidence review happens with every feature, not as a one-time audit.
Feature control as a product, not a config file
Every feature — Timetable, Fee Management, Watchline, WhatsApp credits — is independently switchable per pricing tier from an internal admin panel, with every change audit-logged with before/after state.
AI
13 functions built, 1 switched on
The AI layer runs on Groq's Llama 3.3 70B. Most of it is completed engineering waiting on UI wiring, not unstarted work — a meaningfully smaller lift than building from scratch.
Doubt-answering
LiveRuns on Groq's Llama 3.3 70B. Generates a first-pass answer in the Teacher Doubt Center before a teacher replies.
Daily Knowledge generation
Built, dormantAI-curated article and quiz content for students.
Lesson planning
Built, dormantDrafts lesson plans for teachers from syllabus data.
School health reports
Built, dormantComposite analysis across attendance, marks, tasks and fees.
Weekly AI test generation
Built, dormantA scheduled job already generates an AI weekly MCQ test every Sunday — the student-facing screen to take it isn’t live yet.
PDF syllabus extraction
Built, dormantExtracts structured curriculum data from uploaded PDFs.
Doubt-pattern analysis
Built, dormantSurfaces recurring student confusion at the class level.
General AI chat assistant
Built, dormantA general-purpose assistant across the platform.
Honest roadmap
What’s next, and why it’s a smaller lift than it looks
Close to shipping — integration remaining
- Online fee payments (Cashfree) — schema and transaction bookkeeping done; live API connection remaining
- WhatsApp fee reminders (Meta Business API) — schema and config UI done; live sending not yet connected
- AI Daily Knowledge, homework suggestions, lesson planning, school health reports — functions written and tested; need wiring into screens
Designed, not yet started
- School Health Score — single composite wellness metric
- Syllabus Predictor — AI-forecasted completion date per subject
- Anonymous Class Pulse — anonymous student feedback loop
- Printable report cards
In progress / needs finishing
- Persistent login for students and parents
- Surfacing the already-built Weekly AI Test to students
- Teacher performance analytics
The pattern across nearly every “not yet done” item: the hard part — data model, backend logic, AI functions — is already built. What’s outstanding is largely UI wiring and third-party API connections.
Team
Who's building this
D Sumanth Chandra
Chief Executive Officer
P Manikanta Kowsik
Chief Product Officer
Ch Vamsi Krishna
Chief Technology Officer
Get in touch
Request the deck or a walkthrough
Tell us a bit about yourself and we'll follow up directly.