The Continuing Education Programme (CEP) at the Indian Institute of Technology Delhi (IIT Delhi) has announced the first batch of its Executive Certificate in Agentic AI Systems (Foundation to Deployment), a six-month course aimed at professionals seeking to develop and deploy autonomous artificial intelligence systems for enterprise use.
The programme moves beyond conventional generative AI, which primarily produces text, images and other content, to examine systems that can plan tasks, use external tools, retrieve information and coordinate actions with limited human intervention.
The curriculum covers large language models (LLMs), autonomous AI agents, Retrieval-Augmented Generation (RAG), multi-agent orchestration and production deployment. It also includes AI Trust, Risk and Security Management (AI TRiSM), addressing issues such as model reliability, data protection, explainability and governance.
The course will run online from January 9 to June 19, 2027. The application deadline is October 30, 2026.
Course covers LangGraph, CrewAI and AI security
The curriculum includes eight core modules, according to the institute’s announcement, covering agent design, prompt engineering, enterprise knowledge systems, multi-agent workflows, AI risk management and deployment.
Participants will work with frameworks and tools including LangGraph, CrewAI, AutoGen and the Model Context Protocol (MCP). The programme also covers RAG, which allows AI applications to retrieve information from external documents and databases when generating responses.
Practical exercises will take participants from building individual assistants to developing systems that coordinate multiple agents. Projects listed on the programme website include enterprise document question-answering, natural-language-to-SQL applications and collaborative AI workflows for research and customer support.
The course will conclude with a capstone project in which participants build an end-to-end agentic AI application incorporating AI TRiSM principles.
Professor Sandeep Kumar, Associate Professor in the Department of Electrical Engineering at IIT Delhi’s School of Artificial Intelligence, said: “The challenge is no longer simply to make AI generate better answers. It is to build systems that can take action, use information responsibly and remain dependable when conditions change. We want participants to understand those choices, test their assumptions and develop applications that organisations can confidently use in practice.”
Programme fee, eligibility and assessment
The course is open to graduates from science, engineering and commerce disciplines. The institute recommends at least one year of professional experience, although this is not mandatory. Selection involves an application review and a faculty interview.
The programme fee is Rs 1.45 lakh, excluding 18% GST, according to the official website. Classes are scheduled for Saturdays from 10 am to 1 pm, subject to change. An optional campus immersion is also listed.
Assessment will include examinations, assignments, quizzes and the capstone project. Participants must secure at least 50% in the evaluation and maintain 70% overall attendance to qualify for the Certificate of Successful Completion. Those who meet the attendance requirement but do not reach the minimum evaluation score may receive a Certificate of Participation.
Only electronic certificates will be issued by CEP, IIT Delhi.
Enterprise adoption raises demand for AI deployment skills
The course comes as companies move from experimenting with generative AI to deploying AI agents in business operations. These systems can carry out multi-step tasks, but their use introduces additional concerns around permissions, data access, monitoring, reliability and accountability.
A McKinsey global survey published in August 2026 found that 40% of respondents from organisations with annual revenues exceeding US$1 billion reported scaling AI agents in at least one business function, compared with 27% in the previous year. Adoption among smaller organisations remained at 22%.
Gartner had forecast in August 2025 that 40% of enterprise applications would include task-specific AI agents by the end of 2026, up from less than 5% in 2025. The figure is a forecast, not a confirmed measure of adoption.
Alongside building and coordinating agents, participants will study security safeguards, performance monitoring, cost management and methods to evaluate whether an application is reliable enough for operational use.