Muhammad Aqib User

I’m a Computer Science student specializing in backend architecture, AI integration, and system security. I build asynchronous REST APIs, validate data workflows for Large Language Models, and focus on secure, resilient agentic AI systems.

In my recent role as a Teaching Assistant, I supported undergraduate Multivariable Calculus by designing and grading assignments, quizzes, and projects, then providing academic support to improve student understanding and performance.

My projects reflect my interests in dependable LLM tooling: I built an asynchronous FastAPI Zero-Trust LLM security gateway to block prompt injections, and I developed an agentic Financial Advisory Agent with LangChain and Google Gemini using function calling and strict step-by-step reasoning. I’m also working on evaluating LLM safety guardrails for Roman Urdu adversarial attacks to address vulnerabilities where English-centric filters may fail.