Python Backend Developer · Generative AI Applications

I'm a Computer Science graduate who builds backend systems and AI-powered applications. I like working on the full path from data to API: designing REST services, connecting them to databases, and integrating language models and machine learning where they solve a real problem.

I learn by building, and the projects below are where I put that into practice.

• Backend engineering: REST APIs and service design with Python and FastAPI
• Data layer: relational schemas and database integration with MySQL, PostgreSQL, and SQLAlchemy
• Generative AI: retrieval-augmented generation (RAG) pipelines grounded in real documents
• Applied ML: training, evaluating, and explaining models for practical use cases

A retrieval-augmented QA service for retail documents. Uploaded files are chunked, embedded, and indexed in FAISS; at query time the most relevant passages are retrieved and passed to Gemini, so answers stay grounded in the source text. An optional research-agent mode (CrewAI, Groq, DuckDuckGo) handles questions the documents can't answer. The service is containerised with Docker and built through a GitHub Actions CI/CD pipeline. I replaced local Hugging Face embeddings with API-based embeddings to keep the container image lightweight.

Python FastAPI LangChain FAISS Google Gemini API CrewAI Docker GitHub Actions

A churn analytics platform: an XGBoost model served through a FastAPI backend, with a React + TypeScript interface for running predictions. SHAP explanations show which features drive each prediction, so results are interpretable rather than a bare probability. Containerised with Docker.

Python Scikit-learn XGBoost SHAP FastAPI React TypeScript Docker

• Deepening my FastAPI knowledge, including API authentication fundamentals
• SQL and ORM patterns with SQLAlchemy
• Object-oriented programming in Java
• Building and improving production-minded Generative AI applications

Learning by building. Improving with every project.