AI Engineer & Data Scientist building production-ready agentic pipelines, RAG architectures, MLOps workflows, and LLM-powered products. Kaggle Dataset Grandmaster ranked #26 globally — achieved in just 11 months.
End-to-end AI engineering — from first principles to production deployment.
I'm a Data Scientist & AI Engineer specializing in intelligent systems that actually ship. From agentic pipelines with LangGraph and n8n, to fine-tuned transformers and robust MLOps workflows — I deliver complete solutions.
Currently in my 8th semester at GC University Lahore (BSCS), I've completed 2 ML internships and built 11+ projects spanning agentic AI, RAG systems, computer vision, NLP, and workflow automation.
My philosophy: AI should solve real problems. Every project I build has a live demo, clean code, and a focus on measurable business impact — exactly what modern clients need.
Battle-tested tools across the full AI engineering pipeline.
Production-grade AI systems with live demos and measurable impact.
Autonomous voice agent handling every clinic call — zero human required. ElevenLabs for natural voice, n8n orchestration, Groq LLM for reasoning, and Google Sheets for real-time capture. 100% automated, 24/7 availability.
Production-ready AI health assistant connecting WhatsApp + Airtable + n8n. Instant patient recognition, smart ticket logging, and RAG-based secure record retrieval — 24/7 intelligent support with 99%+ accuracy for clinics.
Production-ready multi-agent research system with LangGraph orchestration, FastAPI backend, Docker deployment, and real-time search capabilities.
Production-grade conversational AI with semantic long-term memory, FAISS vector search, and PostgreSQL-backed knowledge base for grounded answers.
Autonomous personal assistant in n8n powered by Ollama (local LLM) with persistent memory — manages Gmail, Calendar, Tasks, Expenses & Docs in one interface.
Zero-intervention AI that reviews every GitHub PR the moment it opens — posting senior-dev-level feedback as a comment with auto-labeling and coding standards checks.
Complex LangGraph pipeline with router, research, orchestrator, fan-out worker, and reducer nodes for best-in-class long-form blog generation.
Chat with any YouTube video by pasting its URL. Ask AI to extract insights, summarize content, or answer specific questions from hours of video.
Production-ready medical imaging classification using VGG16 transfer learning with full MLOps: DVC pipelines, MLflow tracking, Docker, and CI/CD deployment.
End-to-end ML project with deep EDA, feature engineering, XGBoost & LightGBM models, and a business-decision report for telecom retention strategy.
Fine-tuned DistilBERT on IMDB reviews for sentiment analysis. Deployed on HuggingFace Hub with a Streamlit inference interface for live predictions.
Fine-tuned BioMistral-7B on 10K medical QA samples using QLoRA (4-bit + LoRA). Achieved Perplexity 6.19, ROUGE-1 0.27, BERTScore 0.78 — production-ready biomedical QA on a single GPU with only 0.1% parameters trained.
Verified credentials from leading platforms and institutions.
Open to freelance projects, AI roles, and automation solutions.
Whether you need an agentic AI system, RAG pipeline, automation workflow, or end-to-end MLOps solution — I'll build it with production quality and measurable results.