Summary
Asim Sultan is a Senior Machine Learning Engineer with 8 years of experience building and deploying production-grade AI systems, specializing in LLM fine-tuning, RAG pipelines, and multi-docker GCP deployments. He has hands-on experience fine-tuning open-source models (Llama2/3, Mistral) using PEFT and QLoRA, and operationalizing them with vector DBs like Pinecone/FAISS and orchestration tools such as LangChain and LlamaIndex. Beyond engineering, he teaches advanced Generative AI and multi-modal systems at MIT, where he mentors students on attention architectures, retrieval fidelity, and hallucination mitigation. His background spans end-to-end ML workโfeature engineering, data pipelines, model quantization, and CI/CDโapplied across startups, enterprise, and academic settings. Notably, he emphasizes building internally fine-tuned models rather than relying on third-party LLMs, demonstrating a focus on control, privacy, and reproducibility. Based in Toronto, he blends deep research knowledge with pragmatic deployment skills to turn cutting-edge AI into measurable business impact.
8 years of coding experience
8 years of employment as a software developer
Certificate Program, Practical Data Science and Machine Learning, Certificate Program, Practical Data Science and Machine Learning at Ryerson University - G. Raymond Chang School of Continuing Education
Bachelor's degree, Computer Software Engineering, Bachelor's degree, Computer Software Engineering at International Islamic University, Islamabad
English, Urdu, Pashto