John Alpha

xai-compliance-framework

Overview An AI agent system that performs tasks (search, summarize, advise) while enforcing policy-based guardrails, auditability, and explainability.

Multilingual African Language Sentiment AI Overview A multilingual sentiment analysis platform supporting Shona, Ndebele, and Zulu, designed for social media and customer feedback analysis in African contexts.

Explainable Credit Decision Engine (XAI). A regulatory-compliant credit decision system where every prediction is fully explainable, auditable, and fair. Designed for banking and regulated environments.

An end-to-end MLOps pipeline for training, deploying, monitoring, and retraining financial risk models using AWS-native services. This project highlights production ML engineering, not just model building.

An enterprise-grade Generative AI knowledge copilot that answers user questions over internal documents (policies, SOPs, contracts) using Retrieval-Augmented Generation (RAG) on AWS. This project demonstrates LLM orchestration, secure retrieval, governance, and explainability in production environments.