https://github.com/akinyeraakintunde/enterprise-risk-intelligence-engine

Python-based enterprise risk intelligence engine for log analysis, anomaly detection, KRI-driven scoring, and automated narrative risk reporting. Designed for SMEs, cybersecurity teams, and governance functions. Built as applied AI technical evidence for the UK Global Talent route. Led by Ibrahim Akinyera.
https://github.com/akinyeraakintunde/enterprise-risk-intelligence-engine

anomaly-detection cybersecurity governance log-analysis machine-learning python risk-analytics risk-scoring tech-nation-evidence

Last synced: 3 months ago
JSON representation

Python-based enterprise risk intelligence engine for log analysis, anomaly detection, KRI-driven scoring, and automated narrative risk reporting. Designed for SMEs, cybersecurity teams, and governance functions. Built as applied AI technical evidence for the UK Global Talent route. Led by Ibrahim Akinyera.

• Host: GitHub
• URL: https://github.com/akinyeraakintunde/enterprise-risk-intelligence-engine
• Owner: akinyeraakintunde
• License: mit
• Created: 2025-11-17T22:51:08.000Z (7 months ago)
• Default Branch: main
• Last Pushed: 2025-12-24T04:25:48.000Z (6 months ago)
• Last Synced: 2025-12-25T17:04:00.073Z (6 months ago)
• Topics: anomaly-detection, cybersecurity, governance, log-analysis, machine-learning, python, risk-analytics, risk-scoring, tech-nation-evidence
• Language: Python
• Homepage:
• Size: 2.8 MB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Enterprise Risk Intelligence Engine  

**Live AI Risk Scoring & Narrative Reporting Platform**

🚀 **Live Demo:**  

👉 https://enterprise-risk-intelligence-engine.onrender.com

---

## Overview

The **Enterprise Risk Intelligence Engine** is an applied AI system that transforms raw operational logs and event data into **actionable enterprise risk insights**.

It demonstrates how organisations can move from fragmented operational data to **quantified, explainable, and decision-ready risk intelligence**, delivered through a live, production-deployed interface.

This project focuses on **real-world enterprise risk workflows**, not academic modelling.

---

## What the System Does

**Input**  

📂 Upload a CSV file containing operational events, incidents, or logs.

**Processing Pipeline**

1. Data validation and cleansing  

2. Feature aggregation and enrichment  

3. Key Risk Indicator (KRI) computation  

4. Anomaly estimation  

5. Composite enterprise risk scoring  

6. Automated narrative risk reporting  

**Output**

- 🔢 Overall Risk Score (0–100)  

- 🚦 Risk Classification (Low / Medium / High)  

- 📊 Key Risk Indicators (KRIs)  

- 🧠 Explainable drivers of risk  

- 📝 Plain-English executive summary  

- ⬇️ Downloadable scored dataset  

---

## Live Demo Walkthrough

1. Open the live demo link  

2. Upload a CSV file  

3. View:

   - computed KRIs  

   - enterprise risk score  

   - explainable narrative summary  

4. Download the scored output  

No login. No setup.

---

## Why This Matters

Most risk tools either:

- present dashboards without explanation, or  

- generate black-box scores that cannot be defended to auditors or leadership.

This engine prioritises:

- **Explainability**

- **Governance-aligned metrics**

- **Audit-friendly outputs**

- **Executive-ready narratives**

It mirrors how AI is actually adopted inside regulated enterprise environments.

---

## Use Cases

- Enterprise Risk Management (ERM)  

- Internal Audit & Assurance  

- Operational Resilience  

- Compliance & Controls  

- Cyber / IT Risk Analytics  

---

## Repository Structure

.

├── data/               # Sample and reference datasets

├── diagrams/           # Architecture and system diagrams

├── docs/               # Design documentation

├── examples/           # Example inputs

├── outputs/            # Generated outputs and reports

├── src/                # Core risk logic and analytics

├── streamlit_app.py    # Live demo application

├── requirements.txt    # Dependencies

├── TECH_NATION_EVIDENCE.md

└── README.md

---

## Tech Stack

- Python  

- Pandas / NumPy  

- Streamlit (live UI)  

- Statistical & rule-based risk modelling  

- Explainable narrative generation  

- Render (production deployment)  

---

## Design Philosophy

- Explainability over black-box accuracy  

- Enterprise-aligned risk metrics  

- Human-readable reporting  

- Production realism (live deployment, real inputs)  

---

## Limitations (Intentional)

- Heuristic and statistical methods (no deep learning yet)  

- Demo-scale datasets  

- Single-tenant deployment  

These choices prioritise clarity, governance, and interpretability.

---

## Planned Enhancements

- Machine learning–based anomaly detection  

- Time-series risk trend analysis  

- PDF executive risk reports  

- API-first FastAPI version  

- Expanded CI/CD test coverage  

---

## Project Ownership & Contributors

This project was led and architected by **Ibrahim Akinyera**, who designed the AI architecture, scoring logic, anomaly detection approach, and automated narrative reporting pipeline as part of applied enterprise risk research and UK Global Talent technical evidence.

**Key Contributors:**

- **Ibrahim Akinyera** — AI/ML Lead & Project Architect  

- **Busayo Odukoya** — Technology Risk Expert (domain input, risk frameworks, governance alignment)  

All final technical design decisions, system architecture, and production deployment were led by Ibrahim Akinyera.  

This project forms part of my applied work in:

- AI-driven decision intelligence  

- Enterprise risk analytics  

- Production-ready AI systems  

---

## License

MIT License