kaedax

EdTech · 720-hour cycle

Education software,
built for how India learns.

kaedax builds production education software — AI-learning platforms, classroom and LMS tools, assessment engines — in a fixed 30-day (720-hour) cycle. Curricula align to CBSE's AI skill subject (codes 417/843) and NEP 2020; DPDP-compliant student-data handling, parental-consent flows, vernacular-ready UI, and low-bandwidth-first design are in the spec from day one.

app.gurukul.school/classes/ai-foundations

Active students

1,284

+212 this term

Avg completion

87%

+9% mom

Median quiz

8.4/10

AI Foundations

Need a nudge

23

auto-flagged

StudentClassProgressLast quizStatus
Aarav SharmaX-B92%9/10 on track
Diya KrishnanX-B88%10/10 topper
Mohammed IrfanX-A61%6/10 nudge
Sneha PatilX-C84%8/10 on track
Vikram ReddyX-A43% at risk
dpdp consent ✓ logged · hindi + english live · offline sync 98.4% reconciled
representative build · web console + mobile · anonymized under NDA
Try Gurukul — our live AI-classroom demo 2-min school AI-readiness check

[ §01 ] the problems we're hired for

Where education builds
actually go wrong.

01

Imported LMS platforms don't fit Indian classrooms.

Tools built for US districts assume one device per child, fast wifi, and no board exams. CBSE/ICSE/state-board structures, shared devices, vernacular instruction, and parent visibility are bolt-ons at best — so adoption dies in the staff room.

02

Students are already on AI. The curriculum isn't.

Kids use ChatGPT daily while schools debate banning it. Institutions need structured AI literacy — real syllabus, guardrails, assessment — not a chatbot pasted onto a textbook. That's a pedagogy problem before it's a software problem.

03

Child data is the highest-stakes data there is.

DPDP treats children's data as a special category: verifiable parental consent, no tracking-based profiling, strict purpose limitation. Most edtech retrofits this after launch. Retrofitted consent is how edtech companies end up in the news.

04

Connectivity is a feature decision, not a footnote.

Tier-2/3 schools run on shared bandwidth and intermittent power. If the product isn't offline-tolerant with sync reconciliation, half the country can't use it — and the demo that worked in the boardroom fails in the classroom.

[ §02 ] what ships

One cycle.
Production, not prototype.

The shapes of education work that fit a 720-hour cycle. Every build ships with runbooks, monitoring, and 30–60 days of post-launch on-call.

NEP 2020-ready CBSE AI 417/843 DPDP · child data Parental consent Vernacular-ready Low-bandwidth first Teacher-in-the-loop AI
[01]

AI-learning platforms

Structured AI curriculum products like Gurukul: courses, lessons, quizzes with instant feedback, progress tracking, breakthrough feeds — with eval-gated AI tutors that stay on-syllabus and age-appropriate.

[02]

Classroom & LMS builds

Assignments, grading workflows, announcements, attendance, parent visibility — the Google Classroom shape, adapted to Indian board structures and teacher workflows, in English plus vernacular.

[03]

Assessment engines

Item banks, auto-graded quizzes, rubric-based subjective grading with AI first-pass and teacher sign-off, board-pattern mock tests, and integrity posture appropriate to the stakes.

[04]

Institution ERP & admissions

Admissions funnels, fee management, timetabling, and reporting consoles for school networks and colleges — integrated with whatever ERP the institution already runs.

[ §03 ] proof

all case studies →

Our students were already on ChatGPT; our teachers were a year behind. Gurukul gave us a real AI syllabus, guardrails, and a teacher dashboard — live across 14 schools in one term.

Arun Nair

Arun Nair

Director · 14-school K-12 network, Kochi

[ §04 ] questions

Education founders
ask us first.

01

Is the AI curriculum aligned to CBSE and NEP 2020?

+

Yes. CBSE offers AI as a skill subject — code 417 for Classes 9–10 and 843 for senior secondary — and NEP 2020 expects coding and AI exposure from Class 6. Gurukul's course structure maps to these frameworks, so AI classes slot into the timetable as a recognised subject rather than an extracurricular workshop.

02

How much does edtech software development cost in India?

+

Agency quotes for an LMS or learning platform MVP typically run ₹40 lakh–₹2 crore over 4–8 months. kaedax prices per fixed 720-hour cycle, shared on the scope call — fixed scope, fixed timeline, no overrun billing.

03

Can you build AI tutors that are safe for school students?

+

Yes — that's the core discipline. AI tutors ship behind eval harnesses that test for age-appropriateness, syllabus adherence, and refusal of off-topic or unsafe requests, with teacher-visible logs. Try Gurukul, our live demo: the AI features are guardrailed by design, not by policy document.

04

How do you handle student data under DPDP?

+

Children's data gets the strict reading: verifiable parental consent flows, no behavioural profiling, purpose-limited collection, and data residency in India. Consent state is modelled in the database from day one, not added before a compliance audit.

05

Do you integrate with existing school ERPs and tools?

+

Yes — admissions ERPs, fee systems, Google Workspace for Education, and board-specific reporting formats. Integration chaos is normal in institutions; adapters and sync reconciliation are scoped into the cycle, not discovered after it.

06

Can a learning platform really ship in 30 days?

+

A focused one, yes. Gurukul — the AI-classroom demo on this page, with courses, quizzes, progress tracking, and a teacher dashboard — represents a single-cycle build. Multi-tenant rollout across a school network is typically a second cycle.

Fits a cycle,
or we say so.

A 15-minute scope call with a kaedax founder and the engineer who'd lead your build. We either fit your education product into 720 hours, or we tell you why not — same call.