Mapomo

Political money
is invisible.

We taught machines
to see its traces.

Mapomo MAPPING POLITICAL MONEY

Open-source AI that estimates campaign spending in African elections, from the signals campaigns can't hide.

SCROLL
01 · THE PROBLEM

Campaign spending across Africa is largely undeclared, uncapped or unchecked.

1 / 51

Of 51 African countries, exactly one publishes political donations online.

TRANSPARENCY INTERNATIONAL · 2025

Laws exist. Instruments don't.

A law without an instrument is a poem.

We're building the instrument.

02 · THE INSIGHT
MONEY HIDES. ITS TRACES DON'T.
BILLBOARD₣ · rental price
RALLY₣ · cost per attendee
RADIO₣ · rate card
SPONSORED POST₣ · ad library

A billboard has a rental price. A rally has a cost per person. A radio spot has a rate card. A sponsored post sits in an ad library.

Count the traces.

Know the costs.

Estimate the money.

03 · HOW IT WORKS
01 OBSERVE

Public signals: online ads, billboards, rallies, radio.

02 DETECT

AI models count and classify, including African-language speech recognition.

×
03 ESTIMATE

Counts × real unit costs, calibrated by researchers on the ground.

04 COMPARE

Estimates set against what campaigns officially declare.

ESTIMATED
DECLARED
THE GAP

The gap is the signal.

Mapomo publishes estimates with full methodology, not verdicts. Humans verify; AI counts.

04 · THE TOOLS
IN BUILD · FIRST DEPLOYMENTS 2026 TO 2027

One suite. Four ways to see the money.

IN BUILD

Ad Spend Tracker

Measures online political ad spending, platform by platform.

IN BUILD

Campaign Visibility Index

Converts ground presence (billboards, rallies, posters) into money via real unit costs.

IN BUILD

Money Flow Mapper

Maps the networks behind political financing.

IN BUILD

Local Language AI

Speech recognition and analysis for Wolof, Pulaar and Hausa radio.

05 · THE LANGUAGES

African elections happen in African languages.

On radio in Wolof. In rallies in Hausa. In songs in Pulaar. Tools that only read English miss the story.

RADIO · WOLOF · LIVE TRANSCRIPTION

« …tey ci rajo bi, ñu ngi wax ci kampaañ bi ci diiwaan bi… »

DETECTED · CAMPAIGN MENTION (ILLUSTRATIVE)

Mapomo listens in the languages democracy actually speaks.

06 · WHERE & WHEN

Six countries.
Built for a continent.

SENEGAL NIGERIA GHANA KENYA SOUTH AFRICA DR CONGO

Built in Dakar. Tested where it's hardest. Made for a continent.

SENEGAL
NIGERIA
GHANA
KENYA
SOUTH AFRICA
DR CONGO
NOV 2026 · SOUTH AFRICA
Municipal elections
Remote data run
FEB 2027 · NIGERIA
General elections
First full deployment
AUG 2027 · KENYA
General elections
2027 · SENEGAL
Field validation campaign
07 · THE LAB

28 African experts. 5 days. Saly, Senegal, September 2026.

Journalists, data scientists, auditors, election observers and engineers put the tools and the method to the test, so Mapomo is built by the people who will use it.

THE PARTICIPANTS · 28 EACH CONTRIBUTED IN A PERSONAL CAPACITY
  • Ikechukwu Chukwu
    Ikechukwu Chukwu
    Election observation methods
    NIGERIA
  • Mamadou Diagne
    Mamadou Diagne
    Public data, Wolof speech processing
    SENEGAL
  • Habib Diallo
    Habib Diallo
    Financial investigation
    SENEGAL
  • Ibrahima Diallo
    Ibrahima Diallo
    Applied AI engineering
    SENEGAL
  • Cathy Diouf
    Cathy Diouf
    Data quality control
    SENEGAL
  • Babacar Guèye Diop
    Babacar Guèye Diop
    Country research, campaign prices
    SENEGAL
  • Ayoba Faye
    Ayoba Faye
    Fact-checking and field verification
    SENEGAL
  • Fatou Ndiaye Faye
    Fatou Ndiaye Faye
    Procurement and price comparison
    SENEGAL
  • Seynabou Faye
    Applied AI engineering
    SENEGAL
  • Nafissatou Goudiaby
    Data quality and price series
    SENEGAL
  • Mame Bineta Gueye
    Supplier pricing and collection
    SENEGAL
  • Dr Modou Gueye
    Dr Modou Gueye
    Extrapolation and uncertainty
    SENEGAL
  • Mamoune Ka
    Campaign management and media buying
    SENEGAL
  • Dieynaba Leurs
    Dieynaba Leurs
    Event logistics costs
    SENEGAL
  • Djim Momar Lo
    Djim Momar Lo
    Applied AI engineering
    SENEGAL
  • Arame Sarr Mbaye
    Media advertising market
    SENEGAL
  • Alioune Badara Mbengue
    Alioune Badara Mbengue
    Project lead
    SENEGAL
  • Inssa Toure Mountakha
    Inssa Toure Mountakha
    Applied AI engineering
    SENEGAL
  • Ruth Kutemba Ndongo
    Ruth Kutemba Ndongo
    Country research, campaign prices
    DR CONGO
  • Pape Gorgui Ndoye
    Pape Gorgui Ndoye
    Lab operations
    SENEGAL
  • Malala Raonitsoa
    Malala Raonitsoa
    Campaign spending estimation
    MADAGASCAR
  • Athandiwe Saba
    Athandiwe Saba
    Investigative data journalism, OSINT
    SOUTH AFRICA
  • Stefan Sande
    Stefan Sande
    Election observation methods
    NIGERIA
  • Babacar Sarr
    Babacar Sarr
    Audit and documentary evidence
    SENEGAL
  • Khady Seck
    Khady Seck
    Campaign operations
    SENEGAL
  • Oumou Kalla Tall
    Event services and billing
    SENEGAL
  • Mazamesso Clément Teou
    Mazamesso Clément Teou
    UX and interface design
    SENEGAL
  • Mouhamed Thiam
    Mouhamed Thiam
    Technical architecture and operations
    SENEGAL
08 · OPEN SOURCE

Everything Mapomo builds is published: code, models, methodology.

Because infrastructure for democracy shouldn't have an owner. It should have a community.

GITHUB.COM/MAPOMO-AFRICA→ FOLLOW THE BUILD →
SCHEMAS · THE OUTPUT CONTRACT METHODOLOGY REFERENCE DATA GOVERNANCE
09 · WHO'S BEHIND
ANDAKIA

Mapomo is built by ANDAKIA, an African AI company based in Dakar, makers of production AI in Wolof, Pulaar and Hausa, working with major institutions across the continent.

Supported by a grant from the Open Society Foundations.

Work with us.

Experts, researchers and partners. Mapomo is built with the people who will use it. The first Lab met in Saly in September 2026.

contact@mapomo.org→ FOLLOW THE BUILD