Patrick Attankurugu
Patrick Attankurugu
HomeProjectsBlogTech StackRamah Foundation
CV
Patrick Attankurugu
Patrick Attankurugu

Building AI for financial compliance across Africa

HomeProjectsBlogTech StackRamah Foundation

© 2026 Patrick Attankurugu. All rights reserved.

Projects

Production AI, compliance, and impact systems

A curated list of shipped AI systems across financial compliance, identity verification, computer vision, and social-impact operations.

SENTINEL landing page screenshot
Live

SENTINEL

Autonomous Compliance AI

Live agentic compliance platform: real-time multi-agent investigations, explainable AI, and measured model quality.

Six specialized AI agents orchestrated by LangGraph run end-to-end AML/KYC investigations — sanctions/PEP research, document and graph-based network analysis, explainable risk scoring, and SAR drafting — streamed live, agent-by-agent. Also features an agentic copilot (LLM tool-calling over live data), a model-evaluation dashboard (precision/recall vs labeled ground truth), and semantic case similarity.

PythonLangGraphGeminiFastAPINext.jsRedisPostgreSQL

Live demo — run an agent investigation and ask the copilot

Open project
AfricaPEP landing page screenshot
Open Source

AfricaPEP

Open-Source PEP Screening Platform

Screening interface and open data workflow for Africa-wide PEP checks.

Real-time screening platform for politically exposed persons across all 54 African Union member states. Features fuzzy name matching, graph-based relationship detection, country filtering, and batch processing for KYC/AML workflows.

FastAPINeo4jPostgreSQLNext.jsspaCyDocker

54 AU member states covered

Open projectSource code
Confidential System

KYC/KYB Fraud Risk Engine

Identity & Business Verification

Python
FastAPI
PostgreSQL
ML Models
Enterprise

KYC/KYB Fraud Risk Engine

Identity & Business Verification

Unified KYC/KYB risk pipeline for individual and corporate onboarding.

End-to-end identity and business verification engine covering document checks, liveness, PEP screening, sanctions, and risk scoring for both individual (KYC) and corporate (KYB) onboarding.

PythonFastAPIPostgreSQLML ModelsDocker

Enterprise deployment — client visuals confidential

Private deployment
Confidential System

Deepfake Detection System

Biometric Fraud Prevention

Python
PyTorch
Deep Learning
OpenCV
Enterprise

Deepfake Detection System

Biometric Fraud Prevention

Multi-layer biometric fraud checks running in regulated identity workflows.

Production system detecting AI-generated faces, video replays, and printed photo spoofing in KYC workflows, paired with trained models for intelligent Ghanaian ID capture.

PythonPyTorchDeep LearningOpenCVYOLO

Enterprise deployment — client visuals confidential

Private deployment
Confidential System

AI Credit Scoring Engine

Risk Modeling

Python
scikit-learn
FastAPI
PostgreSQL
Enterprise

AI Credit Scoring Engine

Risk Modeling

Automated decisioning service for credit risk review at financial institutions.

Machine learning credit scoring system that pulls individual and business credit data from XDS Data bureau, applies risk models, and returns automated decisioning for financial institutions.

Pythonscikit-learnFastAPIPostgreSQLXDS Data

Enterprise deployment — client visuals confidential

Private deployment
LexAura landing page screenshot
Live

LexAura

African Regulatory Intelligence

Regulatory Q&A interface for African compliance research.

AI-powered regulatory intelligence platform for African financial compliance. Gives instant, citation-backed answers on KYC, AML, and STR regulations across Nigeria, Ghana, Kenya, South Africa, and FATF standards.

PythonLLMsRAGNext.js

Public app capture

Open project
Ramah Grants landing page screenshot
Impact

Ramah Grants

AI Grants Intelligence

Operational grant pipeline for evidence-aware nonprofit fundraising work.

Funding intelligence platform for Ramah Upliftment Foundation that tracks opportunities, scores grant fit, and helps draft proposal material.

Next.jsDrizzle ORMGroqPostgreSQL

Live Ramah Grants app

Open project