Prepared for the Google Africa Applied AI Lab application · Pitch Deck

NKYMERA

Governed Operational AI for Africa

NKYMERA is building Africa's first governed operational AI platform — translating multilingual requests into structured, authority-controlled workflows for institutions across the continent. Designed to run on Gemini and Vertex AI, anchored in African-language operations, and built with provenance, policy, and human authority at its core.

🌍

Africa-Native Intelligence

Built for African languages, institutions, and operating contexts — not adapted from generic models.

⚖️

Governed by Design

Authority boundaries, refusal-by-default policy, and named human approval built into every workflow.

🔗

Google AI — Powered

Gemini interprets and structures. Vertex AI governs model access. NKYMERA records, routes, and proves.

Enter Deck → Live Demo Platform

Google Africa Applied AI Lab

NKYMERA

Governed operational AI for African institutions — a funding-round pitch centred on business model, TAM/SAM/SOM and demo-led platform proof.

$6–10BTAM hypothesis

NKYMERA-addressable operational AI.

$2–4BSAM hypothesis

Priority markets after Ghana validation.

$30–50MSOM ARR range

Five-year obtainable revenue target.

Future of Work • Ghana Beachhead • Google-native AI01 / 10

Market gap

AI is entering Africa’s operations faster than institutions can govern it.

The missing layer is operational translation: language, policy, authority, workflow state and evidence.

Operational fragmentation

Requests, approvals, assets and evidence remain split across voice, paper, spreadsheets and legacy systems.

Language-to-work gap

African-language intake must become structured, governed work — not only translated text.

Authority risk

Consequential actions require named human approval and refusal-by-default policy boundaries.

Audit deficit

Institutions need evidence of what happened, who approved it, which policy applied and which model acted.

Category thesis: governed operational AI, not generic chatbot automation.02 / 10

Product proof

WorkGraph turns requests into governed execution.

The demo must prove backend state, not UI animation.

Request

Understand

Structure

Verify

Authority

Execute

Evidence

Complete

Multilingual intake

Twi/English field-service request.

Google AI interpretation

Schema-constrained structured work.

Policy and authority

Threshold actions held for named approval.

Provenance

Every consequential event retained.

Launch proof on the separate demo platform.03 / 10

Business model

Enterprise revenue architecture.

NKYMERA monetises mission-critical workflows, not seats alone.

Platform licence

Annual recurring

WorkGraph, governance, evidence, tenancy

Sector modules

Annual recurring

Utilities, public services, logistics, healthcare

AI usage

Usage-based

Gemini / Vertex AI calls under policy

Implementation

Project revenue

Integration, ontology, policy mapping

Managed operations

Annual recurring

Monitoring, model ops, evaluation, SLA

Private deployment

Premium contract

Sovereign / regulated environments

Recurring software + implementation + governed AI usage.04 / 10

Market sizing

TAM / SAM / SOM: explicit, conservative, testable.

Published macro data define the envelope; NKYMERA-specific figures remain planning hypotheses until validated by contracts.

$6–10B / yrTAM

Africa-wide operational AI + workflow + intelligence + integration + managed services.

$2–4B / yrSAM

Priority-country and priority-sector opportunity after Ghana validation.

$30–50M ARRSOM

Obtainable revenue planning range from targeted institutional contracts.

Sources for macro envelope: Mordor Intelligence Africa digital-transformation estimates; GSMA African digital economy reporting. Planning ranges are management hypotheses.

Separate financial model will handle five-year forecast and EBITDA.05 / 10

Expansion strategy

Ghana beachhead. Seven-market expansion.

The map is a scale hypothesis, not a deployment claim. Ghana validates the operating model; each priority market tests a distinct language, institutional and sector context.

Ghana

Beachhead

Nigeria

Scale

Kenya

East Africa

South Africa

Enterprise

Egypt

North Africa

Ethiopia

Population + infrastructure

Angola

Energy + logistics

Priority markets — not current deployments

Ghana first, then tested expansion.06 / 10

Commercial execution

Land with critical workflows, expand into the operating layer.

SOM is built from contracts, not from an arbitrary percentage of Africa.

1 · Paid pilot

One workflow, one authority boundary.

2 · Department

3–5 workflows, shared data model.

3 · Institution

Common WorkGraph and policy layer.

4 · Sector

Reusable module and ontology.

5 · Region

Multi-market operating network.

$30–50M ARRSOM planning range

From targeted institutions and managed operations; validated later in the separate five-year financial forecast.

Commercial motion: land → expand → platform.07 / 10

Competitive position

Africa-native governance is the moat.

Operational intelligence is the category. NKYMERA differentiates through African-language operations, authority-by-design and institutional provenance.

Language-to-work

Operational requests become governed tasks, not translated chat.

Authority by design

The system can stop; automation cannot bypass named approvals.

Provenance as product

Evidence, policy, model, actor and decision history are retained.

Sankofa system logic

Retained evidence and controlled forward action become product architecture.

Palantir validates enterprise operational intelligence as a category; NKYMERA is positioned as Africa-native.

Moat: local operating logic + governance controls + evidence memory.08 / 10

Google AI leverage

Google AI becomes material inside the governed path.

The Lab should accelerate the model-to-operation bridge: Gemini/Vertex AI for interpretation, extraction and evaluation; NKYMERA for policy, authority, workflow and evidence.

Gemini

Multilingual interpretation + structured reasoning.

Vertex AI

Governed model access, evaluation and observability.

Google Cloud

Secure deployment, storage, IAM and telemetry.

NKYMERA

WorkGraph, authority, evidence and institutional memory.

Material Google AI use required in the demo

AI interprets and structures; NKYMERA governs and records.09 / 10

Funding narrative

Use the Lab to convert platform proof into market evidence.

Submission strategy: keep the deck business-led; use the separate demo platform to prove the product.

30 daysDemo-grade WorkGraph

Google AI + authority hold.

60 daysGhana discovery

Lighthouse validation + ACV evidence.

90 daysSector proof

Evaluation report + investor-ready forecast.

Open separate demo platform Restart deck

Ask: technical co-development, Google AI access, evaluation support, GTM validation.10 / 10