Fahad Khalid B.

Stripped TensorFlow, PyTorch, Scikit-learn, Keras, custom model building and computer vision. Those are the lines that would end the technical screen. Replaced with the applied AI he can actually defend, and kept keyword density high since the single profile now reorders by similarity to each job. With 10+ years in C# and .NET, I build enterprise applications on ASP.NET Core, Web API and Entity Framework, deployed on Azure. That covers microservices architecture, SQL Server performance work, RESTful API design, SignalR for real-time features, and full-stack delivery with React and Angular. I work to clean architecture principles, SOLID and DDD, with proper test coverage and CI/CD rather than as an afterthought. ===================================================================== For the past 4 years I’ve worked across Python and applied AI, integrating LLMs into products that are running in production. That means RAG pipelines with citations and confidence gating, multi-agent workflows on LangGraph with human-in-the-loop approval, MCP tool servers, and document classification and extraction using Azure Document Intelligence. Vector search on pgvector and Qdrant, services on FastAPI, deployed in Docker. I work with Claude and OpenAI APIs daily, and Claude Code is my main development tool, so reviewing AI-written code is routine rather than new. What I’m useful for is the join between the two: putting AI into existing enterprise systems without breaking the auth, the permissions or the data boundaries that are already there.