Dates:
October 25 - 27, 2025
Eligibility:
Kifiya Tech staff members.
Project and Submission Requirements
Submission:
-
A functional MVP hosted locally or in a sandbox environment.
-
A GitHub repository or code package.
-
A short demo video (3–5 minutes).
-
A documentation file outlining setup and architecture.
Prizes:
100,000 ETB for the group
Judging Criteria and Metrics
The following metrics will be used to evaluate the MVP submissions. Each project will be scored across three major dimensions: Core Functionality, AI & Technical Quality, and Impact & Innovation.
1. Core Functionality Metrics (40%)
-
Feature Completeness: Percentage of MVP core features implemented (target ≥ 80%).
-
Detection Accuracy (Precision/Recall/F1): AI model correctly flags fraudulent vs. legitimate transactions (target F1 ≥ 0.7).
-
Response Time (API Latency): Average response time to return a fraud score (target ≤ 2 seconds).
-
System Stability: Number of failed calls or crashes during testing (target ≥ 95% success rate).
-
Error Handling & Logging: Graceful error responses and visible logging implemented.
2. AI & Technical Quality Metrics (30%)
-
Model Explainability: Ability of the AI model to explain why a transaction was flagged.
-
Model Efficiency: Performance of the model in terms of computation and runtime.
-
Data Pipeline Design: Clean, modular, and scalable data flow from input to output.
-
Integration Readiness: Ease of integrating the MVP into IFI systems (API-ready, documented).
-
Security & Privacy: Proper handling and anonymization of sensitive financial data.
3. Impact, Usability & Innovation Metrics (30%)
-
User Value Proposition: Demonstrates measurable improvement in fraud detection or financial trust.
-
Innovation Beyond MVP: Unique AI features or enhancements added beyond the core specification.
-
Usability & UX: Intuitive design and clarity of the interface or dashboard.
-
Scalability: Potential to scale across different products or market segments.
-
Presentation Quality: Clarity, confidence, and storytelling in the final demo.
4. MVP Readiness Index (Internal Evaluation)
Used by Kifiya’s internal review team to identify projects suitable for further development. Scored across three factors:
-
Deployment Readiness (40%)
-
Business Alignment (30%)
-
Technical Robustness (30%)
Projects scoring ≥ 70% are considered for incubation or integration pilots.
