Intelligence Stack
What SmartKalamu Is
SmartKalamu Limited is a technology holding company that develops and operates artificial intelligence-powered software platforms, applications, and APIs. Our products serve consumers, institutions, and enterprise clients across knowledge, learning, and information services. We are not simply an EdTech company — we are building the curriculum intelligence infrastructure that the next generation of East African learning platforms will run on.
in one page.
The Insight
Kenya's CBC transition is not an EdTech opportunity — it is an infrastructure opportunity. Schools, teachers, and parents are not the constraint. The constraint is the absence of AI-native curriculum intelligence that understands the Kenya National Curriculum as a structured data object.
SmartKalamu has built that infrastructure. Atta ingests. Taxa classifies. Swali generates. Bongo answers. Paa Elimu and Ace Mtihani deliver it to learners and parents via beautifully designed consumer products.
Why Now
Three structural forces are converging simultaneously:
Key Metrics at a Glance
Infrastructure that never existed.
No CBC-native content intelligence
Existing EdTech platforms were built for 8-4-4. Their question banks, taxonomies, and assessment engines do not understand strands, sub-strands, or competency descriptors. They cannot be retooled — they must be rebuilt.
KCSE orphans the 2025–27 cohort
~600,000 students sitting their final KCSE exams between 2025 and 2027 have no premium AI-native study tool built specifically for their curriculum. The incumbents are PDF dumps and WhatsApp groups.
No shared API infrastructure
Every EdTech startup in Kenya is rebuilding the same content ingestion, taxonomy, question generation, and RAG layers independently. There is no shared infrastructure to build on — until now.
The Market Gap
Kenya has over 3 million secondary school students and 8 million primary school students. The premium private school market alone (500–600 schools) represents a KES 2B+ annual addressable opportunity at modest per-student pricing. No company currently offers an AI-native, CBC-aligned, M-Pesa-native learning platform across consumer and institutional channels — with the API infrastructure to support third-party platforms at scale.
One intelligence stack.
Consumer Products
Kenya's premium KCSE exam prep platform. Mwalimu AI tutor with SSE streaming, Daily Challenge engine, Parent Portal (verdict-first architecture), and M-Pesa STK Push billing. Beta: April 2026. Planned wind-down: 2027–2028 as KCSE retires.
CBC-native learning platform for K1–Grade 12. The long-term consumer flagship. Receives the Ace Mtihani parent audience at the KCSE sunset (Jan 2028). Socratic AI tutor, bilingual, full CBC alignment.
API & Data Products
Content ingestion and chunking layer. Processes PDFs and curriculum documents into structured, semantically indexed chunks. Internal first; externally licensable at KES 8K–65K/month.
Kenya curriculum taxonomy and classification API. Maps content to CBC/KCSE nodes (subject → strand → topic → grade). Sub-licensable to third-party EdTech builders at KES 5K–45K/month.
AI-powered question generation. Three question types, four difficulty tiers, Bloom's Taxonomy and DOK on every item. Platform licensing and institution direct tracks.
Headless semantic RAG API. Grounds AI answers in curriculum-verified content via pgvector. Powers Mwalimu AI across both consumer products. Available to third-party platforms via licensed API.
Content in. Intelligence out.
The Intelligence Pipeline
The Compounding Moat
Each stage of the pipeline produces structured data that trains and improves the next stage. Atta chunks become better as taxonomy signals sharpen. Swali questions improve as Bongo grounding signals surface gaps. The longer SmartKalamu operates, the harder this pipeline is to replicate.
A new entrant cannot buy this advantage. They must build it — and SmartKalamu has a 12–18 month lead on the Kenya curriculum data layer alone.
The B2B Opportunity
Every EdTech startup building on Kenya's CBC curriculum faces the same infrastructure problem SmartKalamu has already solved. Atta, Taxa, Swali, and Bongo are available as standalone licensed APIs — meaning SmartKalamu earns revenue from the broader EdTech ecosystem building on its infrastructure.
This is the AWS model applied to Kenyan curriculum intelligence. SmartKalamu benefits whether competitors succeed or fail — they all need the stack.
Technology Stack
with no incumbent.
TAM
SAM
SOM (Year 2)
| Consumer Segment | Total Students | Addressable |
|---|---|---|
| Form 3 & 4 (KCSE) | ~600,000 | ~120,000 |
| Grade 7–9 (CBC) | ~1.2M | ~180,000 |
| Grade 4–6 (CBC) | ~1.8M | ~150,000 |
| Year 1 target penetration | ~7,000 students | |
| B2B Segment | Institutions | Annual Value |
|---|---|---|
| Top-tier private schools | ~150 | KES 157K+/yr |
| Mid-range private schools | ~400 | KES 80K+/yr |
| API integrators (EdTech) | Growing | USD 599–1,500/mo |
| Year 1 target schools | 20 schools | |
Three revenue architectures.
Consumer Subscriptions
| Product | Channel | Price (KES/student/month) | Viable floor |
|---|---|---|---|
| Ace Mtihani | Parent direct | KES 1,200–1,500 | KES 1,200 |
| Ace Mtihani | School subscription | KES 800–1,000 | KES 800 |
| Paa Elimu | Parent direct | KES 950–1,200 | KES 1,200 (at 1K students) |
| Paa Elimu | School subscription | KES 600–800 | KES 800 |
API & Data Products
| Product | Pricing Model | Entry Tier | Scale Tier |
|---|---|---|---|
| Atta | Per-chunk / subscription | KES 8K/mo → 3,000 chunks | KES 65K/mo → 35,000 chunks |
| Taxa | Per-hit / subscription | KES 5K/mo → 5,000 hits | KES 45K/mo → 75,000 hits |
| Swali API | Per-student / institution | KES 38K/term (150 students) | KES 750K/term (2,500 students) |
| Bongo API | Per-student / developer (USD) | USD 59/mo (500 docs) | USD 1,500/mo (unlimited) |
Strong fundamentals.
Consolidated Portfolio P&L
| Product | Year 1 Revenue | Year 2 Revenue | Status |
|---|---|---|---|
| Ace Mtihani | KES 86M | KES 166M | Beta Apr 2026 |
| Paa Elimu | KES 8M | KES 35M | Sprint 1 |
| Bongo + Swali APIs | KES 11M | KES 38M | Production |
| Atta + Taxa | KES 3M | KES 12M | Productising |
| Total Revenue | KES 108M | KES 251M | |
| Total Costs | KES 75M | KES 145M | |
| EBITDA | KES 33M (31%) | KES 106M (42%) |
Unit Economics — Ace Mtihani (Lead Product)
KCSE to CBC — built in.
| Phase | Window | Milestone | Target |
|---|---|---|---|
| Phase 0 · Beta | Apr 2026 | 200–500 Form 4 beta students. Free. Before/after mock score data collected as proof of learning outcome. | Outcome proof |
| Phase 1 · Launch | Term 2 2026 | 2,000 paying B2C students. KES 950 launch offer. WhatsApp referral mechanic seeded through beta parents. | KES 1.9M MRR |
| Phase 2 · Institutional | Term 2–3 2026 | 20 school licences. 3-week free pilot model. Outcome data report as the sales tool. | 20 schools contracted |
| Phase 3 · Scale | 2027 | 18,000 students. 60 school licences. 2026 KCSE outcome data as primary marketing asset. | KES 166M ARR |
| Phase 4 · Transition | Jan 2028 | Ace Mtihani parent audience migrated warm to Paa Elimu. KCSE sunset handled cleanly. KDPA-compliant data export. | Paa Elimu inherits trust |
B2C Acquisition Channels
B2B & API Channels
Pedagogy. All three.
Founder of Kirimon Market Ventures, a Nairobi-based technology holding company with a 14-product portfolio spanning EdTech, FinTech, AgriTech, and LabourTech. Deep expertise in M-Pesa-native product architecture, investor relations, and multi-product portfolio management across East Africa. Based at Kofisi Karen, Nairobi.
Technical co-founder responsible for SmartKalamu's full engineering architecture — from the Express 5/TypeScript API stack and Supabase pgvector infrastructure to Railway Pro deployment in Africa South-1. Leads engineering sprints across all six products in the portfolio.
Domain expert in Kenya's curriculum landscape — CBC, KCSE, and the pedagogical frameworks underpinning both. Responsible for curriculum alignment architecture, content strategy, and ensuring SmartKalamu's AI outputs meet KICD and KNEC standards at every layer of the stack.
Why This Team Wins
The three failure modes for EdTech startups in Kenya are: (1) great technology, wrong curriculum — solved by Lydia at co-founder level; (2) great curriculum, no scalable tech — solved by Silvia at co-founder level; (3) great product, no commercial discipline — solved by Chamia at co-founder level. These are not advisory hires or future recruits. They are at the founding table, day one.
18 months to portfolio scale.
Use of Funds
Target Investors
SmartKalamu is seeking Kenya-fluent investors who understand M-Pesa-native product distribution, the CBC curriculum transition, and the infrastructure opportunity in African EdTech.