Machine Learning Engineer - VannGuard AI
(2025-01)
- Built RAG-powered AI agents using LLaMA, CrewAI, and LangChain to generate creative briefs from scraped country data.
- Automated Python-based web scraping pipelines using Apache Airflow, integrated with dynamic RAG storage and internal APIs for data retrieval.
- Developed and deployed country-specific chatbots using LLMs and FastAPI, integrating with internal APIs and databases.
- Designed a no-code workflow in n8n to automate ML resume screening: Parsed resumes with Groq-powered LLM → structured JSON (name, skills, experience, links, etc.); Created Drive folders per applicant, uploaded documents, and logged details to Google Sheets; Ranked candidates based on job-specific experience and tech-stack match.
Machine Learning Intern - NECOP (NESCOM)
(2024-01 - 2024-12)
- Developed CNN and ViT models for skin lesion classification with GradCAM explainability.
- Integrated LLM-based QA (Groq + Ollama) for medical support chat.
- Deployed Docker-based ViT training pipelines with GPU acceleration.
ML Intern (Part-time) - CogniMindAI
(2023-01 - 2024-12)
- Explored core ML libraries (NumPy, PyTorch, Matplotlib)
- Implemented CLIPSeg (HuggingFace) for zero-shot image segmentation.
- Built OCR pipelines using PyTesseract on scanned PDF records.
- Contributed to Titanic classifier, ride duration predictor, and LLM prototyping.
Quantum Research Intern - AITECH (NCP)
(2023-01 - 2023-12)
- Studied quantum computing concepts: superposition, entanglement, and quantum gates.
- Researched quantum gates, entanglement, and superposition via IBM Qiskit.
- Developed hybrid SVM-QSVM models integrating classical and quantum ML.
Quantum Computing Workshop - EME NUST
(2023-01 - 2023-12)
- Participated in a one-week hands-on workshop on quantum computing.
- Hands-on training on quantum algorithms via MATLAB and Amazon simulators.
QA Intern - Red Buffer Pvt Ltd
(2022-01 - 2022-12)
- Conducted frontend QA testing for real estate tax apps.
- Authored test cases, tracked bugs, and improved frontend quality.
- Transitioned to JavaScript development and practiced LeetCode for problem solving.
