AI Engineer • LLM Systems • Agentic AI • Trustworthy AI • Computer Vision
AI graduate from FAST-NUCES, Islamabad, building reliable AI systems that combine ML, LLMs, multi-agent systems, and computer vision. Interned at Jazz and atomcamp. Focused on Trustworthy AI, Agentic AI, LLM Evaluation, and Explainable AI.
- AI Intern @ Jazz — Built an ML-powered telecom KPI anomaly detection system, selected for production deployment.
- AI Intern @ atomcamp — Developed LLM-based chatbot components, evaluated language models, created NLP educational material.
HalluciLens (under embargo until submission) — a browser-integrated system for real-time hallucination detection and citation verification in LLM responses, combining retrieval-augmented verification, NLI, citation validation, and automated prompt improvement. Code and paper release pending advisor approval post-submission.
Also open-sourcing the rest of my AI portfolio and building deployable AI systems more broadly.
- Airspace Copilot — Agentic AI aviation monitoring platform (CrewAI, FastAPI, Streamlit, n8n, OpenSky Network).
- Agentic Research Assistant — Multi-agent pipeline for paper retrieval, RAG summarization, and visualization-rich PDF reports.
- Text2Image Generator — Containerized Stable Diffusion inference service (FastAPI, Docker, gRPC, Diffusers, Gradio).
- Selenium Phone Recommender — End-to-end recommendation system (Selenium, rule-based NLP, Tkinter, Flask).
Languages: Python • C++ • SQL • JavaScript • HTML/CSS AI/ML: PyTorch • TensorFlow • scikit-learn • Transformers • Diffusers • spaCy • OpenCV • Keras • NLTK Frameworks/Tools: FastAPI • Streamlit • Docker • Git • CrewAI • n8n • Power BI • Jupyter
Trustworthy AI • Hallucination Detection • Agentic AI • Multi-Agent Systems • LLMs • Explainable AI • Computer Vision