Live, application-led fellowship
Zero to deployed. AI user AI builder.
A 10-week live Agentic AI & AI Engineering fellowship for working tech professionals who already code.
Build RAG, AI Agents, MCP & Multi-Agent Systems — and finish with 9 portfolio projects designed to become deployed, documented proof of work.
10 weeks
50 live sessions
9 projects
40 maximum
₹49,999
₹29,999
incl. GST
No EMI No Refund
Apply for the Fellowship~60-second application Reviewed before counselling
RAGGrounded retrieval
AI AgentsTools & memory
MCPServer + client
Multi-AgentOrchestrated systems
Deployed ProductLive proof of work
Fellowship at a glance
The decision, before the deep dive.
Choose your track
Starts September 12, 2026Both weekday and weekend options
Weekday
Monday–Friday 8:00 PM–9:00 PMWeekend
Weekends 10:00 AM–1:00 PMRecordings available for missed sessions.
Built with someone who builds this for real
Prakash Gudipati
CTO @ Idensys ex-Microsoft ex-Dell
- 15 years building production software
- Architected a facial-recognition platform processing 1.3M+ daily transactions
- Microsoft Certified Azure Solutions Architect Expert
What you build toward
- RAG
- AI Agents
- MCP
- Human-in-the-Loop
- Multi-Agent
- Deployed AI Product
9 projects built progressively
Apply for the FellowshipWhy this fellowship exists
Using AI is not building AI.
Using powerful tools is a start. The next step is learning to engineer systems that retrieve, reason, act, recover and ship.
Using AI
- Asking
- Copying
- Pasting
- Prompting
Building AI
- Designing systems
- Writing code
- Solving real problems
- Creating working proof
Lead instructor
Taught by an engineer who ships production software.
Prakash Gudipati
CTO @ Idensys
ex-Microsoft (Azure Storage) ex-Dell
I Build. I Ship. I Teach.The output
A certificate is a claim. A URL is proof.
Project on your laptop
- Runs only for you
- Stops at the demo
- Difficult to prove
- Disappears into a folder
System you've shipped
- Has a live URL
- Has a repository
- Can be demonstrated
- Becomes portfolio evidence
Nine projects. Built to become proof of work.
The curriculum
Seven phases. Fifty live sessions. One build arc.
- 0110 live sessions
Python for AI
Python patterns and project discipline for real AI codebases.
- 025 live sessions
AI & LLM Foundations
LLM APIs, prompt engineering, structured outputs and reusable wrappers.
- 035 live sessions
Data & Embeddings
Text pipelines, embeddings, vector stores and semantic search.
- 047 live sessions
RAG Mastery
Chunking, semantic and hybrid retrieval, metadata, caching and evaluation.
- 0514 live sessions
Deepest phase
Agents & Agentic AI
Tool calling, ReAct, memory, MCP, LangChain, LangGraph, orchestration, guardrails and evaluations.
- 065 live sessions
Vibe-Coding & Shipping
Build with agentic coding tools, package the product and deploy it.
- 074 live sessions
Career Launch
Portfolio review, GitHub profile strategy and technical positioning.
Every phase ends in a commit. Not just notes. Not just another notebook.
The portfolio
Nine systems. Each one takes you further.
- 01
CLI Productivity Tool
Python foundations applied to a useful tool.
- 02
Reusable LLM API Wrappers
Reliable interfaces for model-powered systems.
- 03
Production RAG Pipeline
Retrieval designed, compared and evaluated.
- 04
Tool-Calling Agent
A reasoning loop that can act through tools.
- 05
MCP Server + Client
Your tools exposed and consumed through MCP.
- 06
Stateful Human-in-the-Loop Workflow
State, branching and a human approval gate.
- 07
Multi-Agent System
Specialised agents coordinated as one system.
- 08
Vibe-Coded AI Product
Built with agentic coding tools and shipped.
- 09
Final capstone
Deployed Agentic AI System
RAG + multi-agent architecture + guardrails + evaluations + live URL + 3-minute recorded demo.
How you learn
Learn by building. Ship by learning.
Activity-Based Learning
Build during the session.
Project-Based Learning
Your projects are the assessment.
Domain-Based Learning
Build in a domain you care about.
CLI Tool→LLM Brain→Semantic Memory→Knowledge Base→AI Agents→Multi-Agent System→Deployed AI Product
Foundations
Python Git/GitHub REST APIs JSON LLM APIs Structured outputs
Retrieval
Embeddings ChromaDB Vector stores Semantic retrieval Hybrid retrieval RAG
Agentic systems
Tool calling ReAct Agent memory MCP LangChain LangGraph Human-in-the-loop Multi-agent orchestration Guardrails Evaluations
Understand the patterns. Then use the tools. Pure Python first. Frameworks second.
Who this is for
You don't need to be an AI engineer yet. But you should already know how to build software.
This is for you if…
- You're a developer moving into AI engineering.
- You follow AI but haven't built substantial systems yet.
- You can code but have nothing meaningful deployed in AI.
- You're building your own AI product.
You do not need: prior professional AI engineering experience.
You do need: a coding/software foundation and willingness to build.
This is not for…
- Complete coding beginners
- Certificate collectors
- Passive learners
- People wanting AI theory without engineering
The commercial decision
Ten weeks to build a portfolio of working AI systems.
- 50 instructor-led live sessions
- 9 progressively built projects
- 1:1 portfolio review
- Weekday or weekend track
- Recordings for missed sessions
- Maximum 40 participants
Admission process
Apply first. Decide together.
- 01
Apply
Complete the short application.
- 02
Application Review
The owner reviews your background and goals.
- 03
Qualification / Fit
Qualified applications move forward; not-qualified applications end here.
- 04
Counselling
Discuss the program and confirm fit.
- 05
Enrollment
Qualified applicants can enroll and pay.
An application is not a payment or enrollment commitment.
Apply in approximately 60 seconds
Start with a short application.
Applications are reviewed before a counselling conversation. Applying does not require payment.
10 weeks
50 sessions
9 projects
40 maximum
- 1 About You
- 2 Background
- 3 Goal
Submitting sends these details to the fellowship application intake. No payment is collected here.