09/30/2026 updated
100 % available
AI/ML Engineer | RAG, AI Agents & Computer Vision | Python, FastAPI, n8n
Karachi, Pakistan
Only remote
B.S. Software EngineeringAbout me
AI/ML Engineer building production AI end to end: RAG systems and AI agents that answer from your data, workflow automation (LangChain, n8n), and real-time computer vision. Python/FastAPI backends. Two published papers. I ship grounded, reliable AI with clean code and clear docs.
AI AgentsRAGOpenAILangChainRESTful API Modeling LanguageApplication Programming Interfaces (APIs)Artificial IntelligenceComputer VisionPython (Programming Language)Machine LearningMongoDBNatural Language ProcessingObject DetectionOpenCVTensorFlow
I am an AI/Machine Learning Engineer who builds and ships AI-powered applications end to end, from data and modelling through to deployed, production-ready systems. My focus is reliable, grounded AI that works on real data and in real workflows, not just in demos.
Generative AI & LLMs
I build RAG (Retrieval-Augmented Generation) systems that answer questions from a client's own documents, databases and websites, using proper chunking, embeddings, vector search (FAISS, pgvector) and re-ranking, with guardrails so answers stay grounded and do not hallucinate. I design AI agents and agentic workflows with LangChain, LangGraph and CrewAI, including tool and function calling, structured outputs and human-in-the-loop approval. I work with the OpenAI, Anthropic (Claude) and Google Gemini APIs, plus prompt engineering, embeddings and LLM evaluation.
Machine Learning & Computer Vision
I build machine-learning and deep-learning models with PyTorch, TensorFlow and scikit-learn. My computer-vision work includes real-time object detection and tracking, image classification and video analytics with OpenCV and CNNs, including a system running across 50+ live camera feeds. I also work on anomaly detection and time-series analysis, evaluated with precision, recall, F1 and AUC-ROC.
Backend & APIs
I develop clean backends in Python with FastAPI: REST APIs, microservices, and integrations with third-party APIs and webhooks. I work with PostgreSQL and MongoDB, and containerise and deploy with Docker.
Automation
I build workflow automations with n8n, connecting AI models, APIs, Google Sheets, email and messaging tools, with retries, error handling and logging.
Frontend
I build web front ends with React, Next.js and TypeScript, and have delivered full-stack applications end to end.
Cloud & Tools
AWS, Azure, Vercel, Render, Git and GitHub, CI/CD, and AI coding tools such as Claude Code and Cursor.
Research
I have co-authored two published research papers: one, as lead author, on multi-modal and green computing for computer vision, and one on explainable AI in medical imaging using Grad-CAM and SHAP.
How I work
I care about what makes AI reliable in production: grounded retrieval, evaluation, clean code, cost and latency, and clear documentation. I communicate clearly, work well asynchronously, and hand over projects ready to run.
Experience
Professional experience0 - 5 years
Freelancer since2024
Languages
EnglishFluent
Project history
Designed and built AI applications and data pipelines in Python/FastAPI for real-time workflows, extraction and automated reporting. Delivered a real-time computer-vision system across 50+ live camera feeds, turning raw streams into operational dashboards.
Validated data flows and APIs using Postman on a 3M+ user application, logged 50+ defects and improved stability by approximately 30%.
Built Generative AI solutions including RAG pipelines with retrieval over pgvector and guardrails to improve relevance and reduce hallucinations. Developed multi-agent, tool-using systems with LangGraph and CrewAI and conversational AI using prompt engineering for real business use cases. Stood up LLM evaluation pipelines tracking accuracy and relevance and built end-to-end ML pipelines reaching 90%+ accuracy. Deployed FastAPI microservices on Docker for scalable, repeatable model serving.