← All programmesGo beyond the prompt.
GENERATIVE & AGENTIC AI / SKILLS ADD-ON
Go beyond the prompt.
Build the workflow.
Generative & Agentic AI
Add generative and agentic AI to your existing skill set. Build retrieval, tools and multi-agent workflows, then learn to evaluate and deploy the system around the model.
3-month skills add-on · Python bridge included
THE AGENT WORKSPACEEXPLORE ↓
Shape a useful response.
Practise prompting and output evaluation, then explore when to use retrieval or fine-tuning for the task.
YOU BUILDBusiness prompt library
Illustrative learning workflow · Select a stage to explore
3 months
Programme duration
10
Curriculum modules
Live
Practitioner-led learning
Build
Projects & capstone
A STARTING POINT THAT FITS YOU
Bring your curiosity. Build your capability.
Add practical GenAI, retrieval and agent-building skills to your existing toolkit.
Professionals who want AI agents working in their business
Plan my starting point ↗DS alumni adding the GenAI/agentic layer
Plan my starting point ↗THE CAPABILITIES YOU DEVELOP
Prompt engineeringLLM fine-tuning (LoRA/PEFT)RAG system designMulti-agent orchestrationMCP tool buildingLLM evals & guardrailsAI app deployment
A focused add-on to your existing skills. Compare the four core programmes →
YOUR LEARNING JOURNEY
From model responses to working systems.
Build through prompting, fine-tuning, multimodal AI, retrieval, orchestration and deployment. Keep evaluation connected to the work throughout.
MODULE 01 / 108 live hours
Python for AI (bridge)
WHAT YOU WILL EXPLORE- Functions/classes refresher, APIs & JSON, NumPy/Pandas for AI workflows
- non-coders start here
THE WORK YOU BUILD
API-driven mini-project
MODULE 02 / 1012 live hours
GenAI Fundamentals & Prompting
WHAT YOU WILL EXPLORE- How LLMs work
- GPT/Claude/Gemini/LLaMA
- zero/few-shot, chain-of-thought, role prompting, multi-turn
- output evaluation
THE WORK YOU BUILD
Prompt library for a business function
MODULE 03 / 1012 live hours
LLM Architecture & Fine-Tuning
WHAT YOU WILL EXPLORE- Transformers & attention (conceptual → hands-on)
- when to prompt vs RAG vs fine-tune
- PEFT/LoRA/QLoRA
THE WORK YOU BUILD
Domain LoRA adapter
MODULE 04 / 106 live hours
Multi-Modal AI
WHAT YOU WILL EXPLORE- Text-to-image, image-to-text, speech (Whisper/TTS)
- multi-modal pipelines
THE WORK YOU BUILD
Multi-modal mini-app
MODULE 05 / 1012 live hours
RAG & Vector Databases
WHAT YOU WILL EXPLORE- Embeddings, chunking, vector DBs
- hybrid search, re-ranking (RRF)
- grounded Q&A, hallucination control, retrieval metrics
THE WORK YOU BUILD
RAG system with eval report
MODULE 06 / 1016 live hours
Agentic Frameworks
WHAT YOU WILL EXPLORE- ReAct, tools, memory, planning
- LangChain agents
- LangGraph state machines
- CrewAI/AutoGen
- human-in-the-loop patterns
THE WORK YOU BUILD
Multi-agent workflow app
MODULE 07 / 108 live hours
MCP & Agent Protocols
WHAT YOU WILL EXPLORE- MCP servers/clients, FastMCP, A2A, structured tool contracts
THE WORK YOU BUILD
Custom MCP tool server
MODULE 08 / 108 live hours
No-Code AI Automation
WHAT YOU WILL EXPLORE- Zapier/n8n/Make workflows
- Notion AI, Airtable
- linking AI to business processes
THE WORK YOU BUILD
Automate 3 real workflows
MODULE 09 / 108 live hours
Evals, Observability & Responsible AI
WHAT YOU WILL EXPLORE- LLM-as-judge, tracing, cost control
- bias/hallucination/privacy
- PII redaction, guardrails
THE WORK YOU BUILD
Eval-gated agent release
MODULE 10 / 1010 live hours
Deployment & Capstone
WHAT YOU WILL EXPLORE- FastAPI/Streamlit/Docker, one cloud deploy
- agentic capstone + demo day
THE WORK YOU BUILD
Deployed agent + demo day
YOUR WORKING TOOLKIT
PythonOpenAI/Claude/Gemini APIsHugging FacePyTorchLangChainLangGraphCrewAIAutoGenFAISSPineconeQdrantMCP/FastMCPLangSmithDeepEvalZapiern8nFastAPIDocker
BUILD AS YOU LEARN
A portfolio with a story behind every piece.
Explore the work attached to each module. Follow the progression from your first exercise to the final capstone.
PROJECT MILESTONE 01
Python for AI (bridge)
API-driven mini-project
CURRICULUM PROJECT / LEARNING ARTEFACT- 01Understand
- 02Build
- 03Review
- 04Explain
CONNECT THE LEARNING TO THE WORK
Functions/classes refresher, APIs & JSON, NumPy/Pandas for AI workflows; non-coders start here
MAKE YOUR WORK EASY TO DISCUSS
What question did you start with? What did you build? Explain one decision, one limitation and what you would improve next.
YOUR PROGRAMME PORTFOLIO
- Prompt library for a business function
- Domain LoRA adapter
- RAG system with eval report
- Multi-agent workflow app
- Custom MCP tool server
- 3 automated business workflows
- Deployed agent with demo day
WHAT THE WORK ADDS UP TO
Learn it. Build it. Explain it.
Work with the model
Explore prompting, fine-tuning and multimodal applications.
Ground the response
Build retrieval pipelines and evaluate their outputs.
Connect the tools
Create multi-agent workflows, MCP servers and business automations.
Evaluate and deploy
Bring tracing, guardrails and a deployed agent together in the capstone.
CAPABILITY DIRECTIONS
GenAI AssociatePrompt EngineerAgentic AI DeveloperAI Engineer (entry)Automation Specialist
These roles describe directions the curriculum supports. They are not a guaranteed job title, employer or salary.
PLAN YOUR STARTING POINT
Turn your interest into a learning plan.
Select what describes you today. Use it to prepare for a conversation about the course, your goals and the time you can commit.
0 / 4
Select the statements that fit. You can still enquire with any number selected.
This is a three-month skills add-on. A Python bridge is included for non-coders; discuss your preparation needs before joining.
Instructors
Taught by practitioners, not presenters
Both instructors hold full-time senior data science roles. You are learning from people who ship this work every week.
PK
Pankit Kumar
Lead Instructor
Sr. Data Scientist, Parexel (a Goldman Sachs–backed company)
10 years in Data Science & AI, building and shipping production systems in regulated pharma/clinical environments. Freelance trainer at BIA, AnalytixLabs and Scaler — he has taught this material to thousands of working professionals.
Data ScienceMLOpsClient deliveryTeaching
IS
Ishaan Sharma
Co-Instructor — AI / GenAI
Senior Data Scientist, Nuvo AI (Meril Group)
Close to a decade across NLP, computer vision and Generative AI. Microsoft Azure ML Scholar. Works on production AI systems in medical devices and healthcare, and leads the agentic-AI and MCP modules.
GenAINLP / CVAgents & MCPLLMOps
BEFORE YOU ENROL
A clear picture of your next step.
Understand the starting point, learning format and portfolio before choosing your programme.
I'm not a developer — can I still take the Generative & Agentic AI course?
Yes. The course starts with a Python-for-AI bridge module for non-coders. You'll need to be willing to write some code by the end, but you don't need prior programming experience to begin. Analysts and product people do well here.
What's the difference between generative AI and agentic AI, and does this cover both?
Generative AI produces content; agentic AI takes actions to complete tasks using tools and planning. This course covers both — prompting and LLMs on the generative side, and LangGraph/CrewAI/AutoGen plus MCP tool-building on the agentic side.
What is the fee and how long is it?
Three months of live mentor-led sessions in small cohorts. The fee is inclusive of GST and all charges, with 0% EMI available; a counsellor shares it on your call.
Will I learn RAG and fine-tuning specifically?
Yes — both are core modules. You'll build a RAG system with a real evaluation report and a domain LoRA/PEFT adapter, because these are the two most in-demand GenAI skills companies hire for right now.
Is there a job outcome attached?
The course targets GenAI, prompt-engineering, agentic-developer and automation roles, and includes assured job interviews through our hiring network plus a post-course internship. The final offer depends on your interview performance.
Read the full Generative & Agentic AI guide +
Beyond prompting: learning to build with AI, not just use it
Almost everyone can now prompt ChatGPT. That skill is worth close to nothing in the job market precisely because everyone has it. What companies are paying real money for is the ability to build with generative AI — to design a retrieval system over their own documents, orchestrate agents that take actions, wire up tools, and put guardrails and evaluation around it so it's safe to ship. That's the gap this course closes.
We take you from GenAI fundamentals and serious prompt engineering, through LLM architecture and fine-tuning (LoRA/PEFT), into RAG and vector databases, then into the agentic layer — LangGraph, CrewAI, AutoGen — and the emerging MCP tool protocols. It ends where every real project ends: evaluation, observability, responsible AI, and deployment. If you're fuzzy on the difference between generative and agentic AI, our explainer on generative AI vs agentic AI is the best place to start.
You don't have to be a developer (but you will code)
The course opens with a Python-for-AI bridge module for people who aren't already coders, so a motivated analyst or product person can keep up. By the end you'll write real code — but the goal isn't to make you a software engineer; it's to make you someone who can turn an AI idea into a working, evaluated system. That's a rarer and more valuable profile than either a pure prompt-writer or a pure coder.
What you'll build
This is a build-heavy programme. You'll ship a prompt library for a real business function, a domain LoRA adapter, a RAG system with an actual evaluation report, a multi-agent workflow app, a custom MCP tool server, three automated business workflows, and a deployed agent you demo on demo day. Every one of these is a portfolio piece — and "I built and evaluated a RAG system" is one of the most bankable lines you can put on a 2026 CV.
Who it suits and where it leads
It fits developers and analysts with basic Python (the bridge module covers non-coders), professionals who want AI agents actually working inside their business, and data science alumni adding the GenAI and agentic layer. It leads to roles like GenAI associate, prompt engineer, agentic AI developer, entry AI engineer and automation specialist. If you want to take it all the way to shipping AI inside client environments, this is also the natural feeder into our FDE pathways.
CONTINUE EXPLORING
Generative AI vs Agentic AI: What They Are and Why the Difference Matters
Everyone's heard of generative AI. Agentic AI is the next step — and the difference is the whole story of where AI jobs are heading.
Read the guide · 7 min ↗What Is RAG (Retrieval-Augmented Generation)? A Plain-English Guide
It's the technique behind almost every useful company AI assistant — and the single most in-demand GenAI skill. Here's what RAG actually is, without the jargon.
Read the guide · 7 min ↗Forward Deployed Engineer Salary in India: Job-Posting Evidence and Offer Comparison
Two live employer disclosures show why one FDE salary band is misleading. Learn how to compare the role, pay basis and offer components.
Read the guide · 4 min ↗YOUR NEXT STEP / GENERATIVE & AGENTIC AI
Go beyond the prompt.
Build the workflow.
Get the detailed syllabus and talk through your starting point, weekly schedule and the work you want to build.
3 monthsLive practitioner learningFees shared on a call
Compare all programmes →