Enrolling in the Postgraduate Certificate programme at IIT Mandi was one of the most deliberately uncomfortable decisions I've made. I was already running TechTrio Automation, shipping code for government clients, and managing production infrastructure. Adding an intensive academic programme on top felt borderline irresponsible. It turned out to be exactly the right call.
The curriculum covered the foundations I had gaps in: transformer architectures, attention mechanisms, prompt engineering at a systems level, and fine-tuning strategies for domain-specific models. What I appreciated most was the emphasis on applied work — every module tied theory back to real deployments.
The most valuable module was on Retrieval-Augmented Generation (RAG). We built a complete pipeline: document ingestion, chunking strategies, embedding models, vector databases, and LLM orchestration. I immediately saw applications for client work — especially for the knowledge-management systems some of my enterprise clients had been asking about.
Working alongside engineers from top companies across India was its own education. The peer conversations about scaling challenges, model selection trade-offs, and cost optimisation were as valuable as any lecture. I came away with a much clearer model for when to reach for a general-purpose LLM versus fine-tuning versus a RAG pipeline.
If you're a practising engineer wondering whether a structured AI programme is worth it alongside full-time work: it is, provided you have the discipline to stay consistent over a sustained period. The structured curriculum forces you to build foundations you'd otherwise skip in the rush of day-to-day delivery.