On 8 August, HCLTech and the Economic Times ran an eight-hour online AI Masterclass for second-, third- and fourth-year engineering students, covering AI-assisted development, large language models, and AI agents. Around 300 students from 70 colleges - campus ambassadors and top performers from AMPlified - The AI Challenge - took part, with the goal of making them industry-ready in a hiring market that increasingly rewards practical AI skills over classroom theory.
Opening the session, HCLTech's Corporate Vice President and Global Head of Digital Business, Pawan Vadapalli, said skills in the AI era go beyond coding. Students need to learn how to frame problems, understand business context, communicate ideas, and exercise engineering judgment. "AI can make coding faster," he said, "but it does not remove the need for engineering thinking."
From GitHub Copilot to RAG
The curriculum moved progressively from fundamentals to advanced development. Students started with AI-assisted software development using GitHub Copilot, learning how AI can write, explain, and test code - including edge cases - while also understanding why code quality, security, and maintainability matter, not just speed.
Sessions then covered large language models, embeddings, and Retrieval-Augmented Generation. Students explored tokenisation, context windows, hallucinations, document loaders, and text splitting before building knowledge-aware applications. A financial RAG use case demonstrated how enterprise applications can ground LLM responses in proprietary information.
AI agents and cloud deployment
The final modules introduced AI agents, agentic architectures, and cloud deployments, including LangChain, LangGraph, multi-agent systems, and MCP (Model Context Protocol) Training, along with an introduction to AWS Bedrock. Students also explored practical applications such as code-review agents and learned when to use LangChain rather than LangGraph.
A recurring question among students was how to stay relevant as the technology and career landscape evolves. They were encouraged to focus on AI infrastructure - databases, memory, guardrails, and deployment - while building a strong foundation in programming, data structures and algorithms, system architecture, and Generative AI and LLM Courses fundamentals. The message repeated throughout: the best way to learn AI is by building.
Vadapalli stressed that technology will evolve rapidly, but the traits employers value will remain constant: curiosity about what is happening in AI, evidence of having built things independently, agility in responding to challenges, and the passion to keep learning.
Beyond the technical
Students also raised questions about AI and data sovereignty, the sustainability of AI infrastructure, and the environmental impact of data centers. Vadapalli described sustainability as an organisational culture and pointed to HCLTech's focus in that area.
The masterclass concluded with students required to submit two AI projects to qualify for completion certificates - putting the emphasis on learning by building rather than learning by listening. The session was part of the ET Masterclass-HCLTech AI Skills for the Future series for B.E. and B.Tech engineering students.
Why this matters for educators
For faculty and course designers, the session offers a clear template for what industry now expects from engineering graduates: applied projects over theoretical coverage, grounding in LLM fundamentals, and exposure to deployment tools. The requirement to submit two projects as a completion condition is a model worth borrowing - it forces students to demonstrate competence, not just attendance.
Your membership also unlocks:
