Africa's data.
Finally built.

Every major AI in the world was trained on internet data. The internet is approximately 98% non-African. The internet is 98% non-African. That means 98% of African prices, behaviour, context, and reality was never in the training data for any AI. AfriFoundry is building what was left out — verified by humans who were actually there.

500,000+ verified · on the road to 1,000,000

500,000+
Verified datapoints
on the road to 1 million
30+
Markets sampled
15
Sectors covered
47
Counties represented

Where the data lives

Every bubble is a real location. Size and intensity reflect datapoint density — where we've collected the most, and where we're still building. Kenya is the primary market. East Africa is next.

TOTAL MAPPED
500,000+
on the road to 1 million

Hover any bubble to see location details · Scroll to zoom · Data updated as scouts submit

98%

What global AI ignored

The internet is 98% non-African. Every AI — Claude, GPT, Gemini — was trained on that data. Which means African prices, markets, languages, and context were almost entirely absent from every intelligent system on the planet.

2%

What AfriFoundry is building — the 98%

The 98% of African data that global AI never had. Real prices, real markets, real context — collected at ground level by people who were actually there. Verified, structured, permanently owned by AfriFoundry.

Why this is a moat

No competitor can train on data that doesn't exist. We have a 9-month head start building a dataset that will power every AI that wants to serve Africa — including ours.

Nothing junk enters the database.

Every datapoint passes through a 5-stage pipeline. Every point has a location, timestamp, source, and confidence score. Nothing estimated. Nothing scraped from Wikipedia.

01
Collected
Scout walks the market, records price, location, and timestamp
02
Normalised
Cleaned and standardised into the three-table schema
03
Validated
Confidence-scored — below 0.65 goes to manual review queue
04
Deduplicated
Cross-checked against existing datapoints by location and product
05
Stored
Tagged across three axes: Geography × Industry × Intelligence Layer

Real data. Real places. Real people.

12 data types across 15 sectors. Every category has a scraper, a field collection protocol, or both.

📦 Market prices (produce, goods, services)
🏗️ Construction costs by county
🚌 Transport fares and routes
💊 Healthcare costs and availability
📱 Telecoms and data pricing
🏠 Rent and real estate by area
👨‍🌾 Farm-gate prices and yields
⚡ Power access and reliability data
💰 Labour rates and wage floors
📋 Licensing and regulatory requirements
🏪 Competitor pricing by sector
🌍 Consumer behaviour patterns

Markets we've walked

Our scouts and founder have personally visited every market below. The data wasn't estimated — it was collected on the ground, by people who were there.

Gikomba
NAIROBI
Wholesale clothing & goods
Kongowea
MOMBASA
Fresh produce
Wakulima
NAIROBI
Produce wholesale
City Market
NAIROBI
Mixed retail
Marikiti
MOMBASA
Fresh produce
Eastleigh
NAIROBI
Clothing & electronics
Toi Market
NAIROBI
Second-hand goods
Kibuye
KISUMU
Fresh produce & fish
Karatina
NYERI
Produce & livestock
Kongowea Phase 2
MOMBASA
Electronics & phones
Likoni
MOMBASA
Mixed retail
Limuru Road
KIAMBU
Farm gate prices

15 sectors. Growing.

Food & AgricultureHealthcareConstruction & Real EstateTransport & LogisticsEducationTechnology & TelecomsFinance & M-PesaRetail & TradeEnergy & PowerWater & SanitationTourism & HospitalityManufacturingMedia & ContentGovernment & RegulationLabour & Employment

See the data in action

Every conversation with AfriFoundry AI draws from this dataset in real time. Ask about prices in your market. See what 500,000+ verified African datapoints feels like.

Talk to AfriFoundry AI →