Investor Overview

Back Ethiopia's
financial rail.

This is not a typical fintech bet. Sheba Credit is the foundational credit infrastructure for a 135-million-person economy — at the precise moment digital rails, regulatory reform, and market demand converge.

Current Round
$500K
Seed Round · 2025–2026
Stage
Pre-Revenue Seed
Use of Funds
Tech + Licence + Ops
Break-Even
Year 2
5-Year Cum. Profit
$5.2M
Founder
Hamid Ali
Market Opportunity

A $5B gap.
Zero private
bureaus.

Ethiopia has under 2% credit bureau coverage — the lowest of any major African economy. Kenya had similar coverage 15 years ago, then M-Pesa and bureau integration drove explosive credit expansion. Ethiopia is at that same inflection point today, with Telebirr's 51M users providing the same data foundation.

Sheba Credit is the first mover building Ethiopia's private bureau — with AI and open banking baked in from day one. The window to set the technical standard is now.

South Africa64%
Kenya52%
Nigeria18%
Ghana12%
Ethiopia (Today)<2%
Sheba Credit's target: 3M+ Ethiopians with credit files by Year 5 — the same penetration Kenya achieved in its first bureau expansion phase.
AI-Powered Engine
The Technology

Machine learning meets
Ethiopian reality.

Sheba Credit doesn't import a Western scoring model and hope it works in Africa. We build AI trained entirely on Ethiopian alternative data — mobile money flows, savings behaviour, agricultural output, and utility payments — making it uniquely accurate for the 98% of the population that traditional models cannot score at all.

Sheba AI Scoring Pipeline — End to End
📡
Alternative Data Ingestion
Telebirr · Utilities · Savings · ECX · Open Banking
5 Sources
🧹
Feature Engineering & Cleaning
Normalisation, outlier removal, seasonal adjustment
Real-time
🧠
ML Ensemble Scoring Engine
XGBoost + Deep Neural Network combined
Ensemble
🌾
Agricultural GNN Layer
PyTorch graph neural net — satellite + market data
GNN
🔍
SHAP/LIME Explainability Layer
Regulator-ready factor breakdown — Amharic + English
NBE-ready
CRaaS API Output
Full credit report delivered in under 200ms
<200ms
Sample Sheba AI Credit Report — Live Output
742

742 / 850 — Good Standing

Loan tier: ETB 15,000–50,000
Risk level: Low · Model confidence: 94.2%

Updated in real-time · Explainable by law

Telebirr Activity
88%
Savings Consistency
74%
Utility Payments
92%
Financial Literacy
61%
Open Banking
54%
🎯 XGBoost + Neural Network Ensemble Core Model

Our engine combines XGBoost gradient boosting with a deep neural network. The ensemble consistently outperforms either model alone — especially for thin-file customers with limited history.

  • Trained exclusively on Ethiopian alternative data from day one
  • Scores users with as little as 3 months of Telebirr transaction history
  • Continuously retrained as real loan outcomes flow back from B2B partners
  • Separate sub-models for urban consumers, rural farmers, and SMEs
🌾 Agricultural Graph Neural Network Unique to Sheba

15M+ Ethiopian smallholder farmers have zero credit history. Our PyTorch GNN maps relationships between farmers, cooperatives, crop types, and ECX market prices to generate agricultural credit scores no traditional bureau can produce.

  • Satellite imagery analysis for crop yield estimation
  • ECX commodity price integration for seasonal income modelling
  • Cooperative network graph — shared repayment track record
  • Weather and climate risk overlays via ACRE Africa data partnership
🔍 SHAP/LIME Explainability NBE Compliant

Every score is fully explainable. Regulators, lenders, and consumers can see exactly which factors drove the decision — a legal requirement under Ethiopia's emerging data protection framework and a key differentiator with institutional clients.

  • Adverse action notices auto-generated in Amharic and English
  • Full factor breakdown visible to consumer in-app
  • Immutable audit trail stored for every score ever issued
🔄 Continuous Learning Loop Self-Improving

Every loan issued using a Sheba score feeds repayment outcomes back into the model. The AI compounds in accuracy over time — creating a data moat no new entrant can quickly replicate.

  • Model drift monitored daily for accuracy and bias
  • Federated learning design protects individual data privacy
  • Each new B2B partner strengthens the model for all others
Real-Time API at Scale <200ms

The CRaaS API returns a full explainable credit report in under 200ms — enabling instant decisions at point of application, mobile loan screen, or bank branch counter.

  • Redis caching + Apache Kafka streaming pipeline
  • 99.9% uptime SLA · scales to 150K+ checks/month
  • Webhook support for real-time default alerts on portfolio
5-Year Financial Projections

Break-even
by Year 2.

Conservative model: $2/check B2B pricing · 15 institutions Year 1 · scaling to 300+ partners and 7.5M annual checks by Year 5.

YearRevenueOPEXProfit/LossCumulative
Y1 2026$210K$277K−$67K−$67K
Y2 2027$500K$320K+$180K+$113K
Y3 2028$1.0M$400K+$600K+$713K
Y4 2029$2.0M$600K+$1.4M+$2.1M
Y5 2030$4.0M$900K+$3.1M+$5.2M

Sensitivity Analysis — Year 1

Conservative (2,500 checks/inst.)
−$142K
Base Case (5,000 checks/inst.)
−$67K
Optimistic (10,000 checks/inst.)
+$83K

Use of Funds

$500K
Seed Round Allocation
Core Technology & AI Models
$200K40%
Cloud & Security Infrastructure
$80K16%
App & Web Development
$70K14%
Operations & Office (Yr 1)
$75K15%
NBE Licensing & Legal
$25K5%
Marketing & Pilot Acquisition
$25K5%
Contingency Buffer
$25K5%
Year 1 Revenue
$210,000
Partner Institutions
15 (Year 1)
App Downloads
50K (Yr 1)
Cash-Flow Positive
Year 2
5-Year Cum. Profit
$5.2M
Year 5 Checks/Year
7.5M
Request Full Investor Pack
Competitive Advantage

Four durable moats.

01

First-Mover in a Nascent Market

No private credit bureau exists in Ethiopia. Sheba sets the technical standards, trust norms, and regulatory precedent. Licensing requirements and compliance costs create a durable barrier once we're established.

02

Alternative Data Moat

By scoring farmers, women, and youth first through Telebirr, cooperative, and utility data, Sheba builds datasets competitors can't quickly replicate. Every new credit file strengthens the model — a self-reinforcing compounding advantage.

03

AI Trained on Ethiopian Reality

Unlike imported scoring systems, Sheba's ML models are trained on local alternative data from day one — uniquely accurate for Ethiopia's informal economy. SHAP/LIME explainability satisfies regulators and builds institutional trust.

04

Powerful Policy Tailwinds

Ethiopia's Vision 2030, Fayda digital ID (17M enrolled), the Ethiopian Securities Exchange, and foreign bank entry all massively raise demand for bureau-grade data. Sheba is the infrastructure layer national policy requires.

5-Year Roadmap

Foundation to East Africa leader.

Year 1
Foundation
Secure NBE licence
Launch Sheba Score + Save
15 institution partners
50K app downloads
Education launch
$210K revenue
Year 2
Scale
Launch Sheba Agri
50 institution partners
100K app users
USSD rural access
ISO 27001 certified
$500K · Cash-flow +
Year 3
Profitability
250K users
100 institution partners
Real-time AI scoring APIs
Djibouti pilot
Education V2 — 5 langs
$1M · Break-even
Year 4
Regional
Enter Somalia + S.Sudan
200+ partners East Africa
500K users
Blockchain dispute layer
Diaspora credit product
$2M revenue
Year 5
Leadership
1M+ users
300+ partners
7.5M checks/year
NBE strategic partner
East Africa standard
$4M · $5.2M cum.
Invest in Sheba Credit

"Back Ethiopia's
financial infrastructure."

The $500K seed round is open. Request the full investor pack — detailed financials, cap table, technical architecture, and regulatory roadmap.

Request Investor Pack ▶ View Live Demo

Headquartered in Addis Ababa, Ethiopia · Seed Round: $500K Open