Emmanuel Nwanguma
ML Engineer·LLM Integration·RAG Pipelines·Production AI
Ready for work
Emmanuel is ready for their next project!
Fine-tuned Microsoft Phi-4 Mini (3.8B) on SEC financial Q&A using QLoRA. Achieved +69% ROUGE-L over base model. Live demo on Hugging Face.
0
34
Production ML system for heart disease risk prediction. 88.5% accuracy. FastAPI REST API, SHAP explainability, MLflow tracking, Docker deployment.
0
34
Production RAG pipeline with FAISS vector search, semantic chunking, and LLM observability. Sub-500ms retrieval on 100K+ documents. FastAPI backend deployed on Render
0
33
Built evalflow — an open-source pytest-style quality gate for LLMs. Catches prompt regressions before production by diffing live outputs against a saved baseline. Live on PyPI. github.com/emartai/evalflow
0
42