Engineer LLM applications, RAG, agents and MCP integrations with the evaluation, guardrails and observability that real users require.
Companies are moving from prompts to systems that retrieve knowledge, call tools and take actions. They need engineers who can measure quality, control cost and stop an agent doing the wrong thing.
Every week has a hands-on lab and an assignment. Open a week to see both.
Control model behaviour
Call two providers through a gateway and stream the responses.
Compare quality, cost and latency for a support use case.
Extract structured fields from dispute emails with schema validation.
Version your prompts and add regression tests.
Embed and index card-scheme rules in pgvector.
Benchmark two embedding models on your own queries.
Ground answers and take actions
Build a cited assistant over regulatory circulars.
Measure retrieval quality before and after reranking.
Build an agent that pulls transaction history and checks rules.
Define tool contracts and handle every failure path.
Expose your own tools through an MCP server.
Add memory to your agent and trace a multi-step run.
Measure, secure and ship
Build a golden dataset and LLM grader for your assistant.
Block merges that drop eval scores below a threshold.
Red-team your agent with prompt injection attempts.
Add guardrails and permission checks on every tool.
Trace every request end to end in Langfuse.
Cut cost per request with caching and model routing.
Agree scope and milestones for your capstone workflow.
Deliver the deployed workflow with tracing and a cost report.
The fast track covers the core in 10 weeks. The standard pace spreads the same content and the capstone over up to 12 weeks.
Plan for 8–10 hrs/week, including the live session.
A cited assistant over card-scheme rules, RBI circulars and internal policies.
Public regulatory circulars and synthetic internal policies
An agent that pulls transaction history, checks rules and drafts an investigation note.
Flagged transactions with merchant and customer history
An agent that matches ledger and bank entries and routes exceptions for human review.
Ledger, bank statement and PSP settlement files
Build a production-style agentic workflow that triages customer disputes, gathers evidence through tools, drafts a resolution and escalates edge cases — with evaluation and guardrails throughout.
Dataset: Customer disputes, transaction logs and scheme rules
Prefer another domain? Choose SaaS, E-commerce, Logistics or Healthcare at the capstone kick-off.
Build LLM features and RAG systems
Design tool-using, auditable agents
Evaluate, monitor and optimise LLM systems
These are the roles this track prepares you for. AI4Impact provides career support; it does not guarantee employment, interviews or a particular salary.
Completion records that you did the work. The professional certificate is awarded only after an independently reviewed capstone and a recorded defence, so it means something to the people hiring.
Add it to any track. Five modules that take you from a finished capstone to real interviews, for ₹24,999.
Each week has short recorded lessons, a hands-on lab and an assignment, plus a live clarification session with your trainer. Plan for 8–10 hours a week, including the live session.
They are part of Programme+: five one-to-one 30-minute sessions you book through your learner dashboard and use for scoping, reviews or career direction.
Fintech is the default. If you prefer another domain, you can choose SaaS, E-commerce, Logistics or Healthcare at the capstone kick-off.
No. AI4Impact provides career support and recommendations, but it does not guarantee employment, interviews or a particular salary. Hiring decisions are made by employers.
Book a free 20-minute call. We look at your background and tell you honestly which track fits, or if none of them do yet.
