You ask ChatGPT for answers.
You don't know how it works.
In AI Residency
Learn what a model is, why it hallucinates and how to make it do what you want.
IIT Patna Vishlesan i-Hub 
3 Days Goa4 Weeks Online20 Nov Stop prompting.
Apply Now
Stop prompting.
Start shipping.
Built for College studentsFresh graduates No coding needed
- Build with AI, for real – 3 days in Goa and 4 weeks online. Build and deploy a live AI product.
- Investor–Founder Mixer – show investors and startup founders what you built, in person in Goa.
- Earn an i-Hub, IIT Patna certificate – a course completion certificate from Vishlesan i-Hub, IIT Patna.
5 seats left
Only 50 seats. Starts 20th Nov 2026. Every application is reviewed personally.
Internship opportunity ₹25,000 free AI credits*
At a glance
Exactly what you get.
3days Residential in Goa from 20 Nov, at Fairfield by Marriott
4weeks Online: 2 live sessions and a 1.5-hr lab every week
1live product Built from Goa to Week 4 and deployed to a real URL
50seats One cohort: you and 49 other builders
8sessions Live online, plus 4 labs — and 3 teaching sessions, a build day and a lab in Goa
You walk out with
- A live AI product on a real URL
- A pitch you've actually delivered
- A founder network
- A certificate from i-Hub, IIT Patna
Taught by engineering leaders from companies building with AI
Is this you? Everyone can prompt.
Everyone can prompt.
Almost nobody can ship.
AI Residency is for you, if…
You ask ChatGPT for answers.
You have ideas.
You can't build them.
In AI Residency
Write a spec, cut it to an MVP and build it with Cursor, Claude Code or Codex.
AI writes the code.
You can't read it.
In AI Residency
Learn where your code runs, how APIs talk, and how to debug what an agent wrote.
Your degree says you studied.
It doesn't say you can build.
In AI Residency
Leave with a live product on a real URL, and show it to investors and founders in Goa.
Not “learn to code with AI.” Learn enough technology to build intelligently in the AI era, and leave with proof.
The AI Residency Journey
3 days in Goa. 4 weeks online.
Phase 01 Goa 3 Days
Experience, connect, build.
Sessions, a full build day and the whole cohort in one room. Arrive curious, leave with a working product, and meet investors and founders at the Investor–Founder Mixer.
Day 01
Arrive & meet
Check in, then the opening session on the AI-native builder, structured cohort introductions and an evening conversation with an AI and product leader.
Day 02
Build day
Take apart a real product, scope a small idea, then build and deploy it to a live URL with your pod. Every pod demos at show and tell.
Day 03
The web & data
How the web and APIs work, databases and system design, and reading AI-generated code. Then an evening lab wiring a front end to a real backend.
Full Goa scheduleMorning to evening, all 3 days
Day 01Arrive & meet
- MiddayArrival & check-in. Welcome lunch and orientation.
- AfternoonOpening: The AI-Native Builder. Why the programme exists, how AI is changing the way products are created, and what every builder should be capable of by the end.
- AfternoonThe cohort in one room. Structured introductions: backgrounds, ambitions and the problem each builder wants to solve.
- EveningInside the AI revolution. A conversation with an AI, product and technology leader on how AI is changing businesses and which skills will matter. Then cohort dinner and networking.
Day 02Build day
- MorningProduct teardown. Take apart a real product the way a product manager would: the user, the problem, the design choices and the flow from first screen to checkout.
- Late morningFrom idea to working product. What the user touches, what happens behind it, where data lives and where AI fits. Then scope a small idea you can build by dinner.
- AfternoonBuild it. Pods build and deploy a real, working product to a live URL with AI tools, with mentors on the floor throughout.
- EveningShow and tell. Every pod demos what it built and names the one thing it couldn't explain. Those gaps become the syllabus. Then cohort dinner.
Day 03The web & data
- MorningThe web & how applications talk. URLs and DNS, HTTP, frontend and backend, and APIs. Case study: trace one order through the platform.
- AfternoonData, architecture & reading code. Databases, SQL vs NoSQL, system design basics and reading AI-generated code. Case study: design the platform's data model.
- EveningHands-on lab. Call real APIs, wire the front end to a backend and database, and read the code the AI generated.
What leaves Goa with you
- A working product on a live URL, built on Day 2
- A data model for the platform, with the reasons behind it
- A pod, a mentor and the gaps you named at show and tell
- The web and data sessions behind you, so Week 1 starts at depth
Phase 02 Online 4 Weeks
Models, product, agents, production.
Week 1 starts the week after you land home: two live sessions and a 1.5-hour lab every week, each taught against the platform you started in Goa.
- Week 01
Models, then control
What an LLM is and why models behave differently, then prompting that holds: structured outputs, context engineering and tool calling.
See sessions & lab
Session 1
The AI landscape & LLMs
- Where AI actually is right now, past the headlines
- Chat vs copilots vs agents
- What an LLM is; models and model families
- Tokens and context windows
- Why models hallucinate and behave differently
- Multimodal models
- Open vs closed models; exploring models on Hugging Face
- Cost, latency and model choice
Session 2
Getting AI to do what you want
- System vs user prompts
- Few-shot examples
- Structured outputs
- Context engineering
- Tool calling
- Streaming
- Debugging AI outputs, and why they failed
- Common prompting mistakes
Lab · 1.5 hrsBuild the platform's product-description generator with progressively better prompts, and keep the ones that hold.
- Week 02Evaluation 1 · 45%
Idea → working product
Turn an idea into a spec and plan before you prompt, then build the storefront as a running app with coding agents like Cursor, Claude Code and Codex.
See sessions & lab
Session 1
Structured AI-assisted development
- Turning an idea into a product specification
- MVP definition: what is in, what waits
- Why blindly vibe coding fails
- Planning before prompting
- Breaking large work into smaller tasks
- A reusable skill library
- Git and GitHub: commits, branches, version history
Session 2
Building with AI coding agents
- Cursor, Claude Code and Codex: what each is good at
- How coding agents work
- Giving agents context, specs and skills
- Reviewing generated code
- Modifying an existing application with an agent
- Debugging generated code
Lab · 1.5 hrsBuild the storefront as a running app: catalogue, cart and orders persisting properly.
Evaluation 1 · 45%The storefront, the data behind it and the AI-assisted workflow that produced it, explained, not just shipped.
- Week 03
Agents, your data & the outside world
From one agent to many, with automations in n8n. Then ground the platform in your own data with RAG and connect it to other systems over MCP.
See sessions & lab
Session 1
Agents → multi-agent systems
- LLM call vs workflow vs agent
- Tools and function calling
- Memory between turns
- Agent loops and human-in-the-loop
- Guardrails, retries and knowing when to stop
- Sequential, parallel and supervisor patterns
- Handoffs and shared state
- Workflow automation with n8n
- Agent cost and caching
Session 2
RAG & MCP
- Embeddings: comparing text by meaning
- Vector databases and similarity search
- Chunking
- The retrieval pipeline, end to end
- RAG vs fine-tuning
- Connecting AI to applications and databases
- Using MCP servers
- A2A, agent-to-agent communication, in brief
Lab · 1.5 hrsA support agent grounded in the store's policies and connected over MCP, with the post-order flow automated in n8n.
- Week 04Final evaluation · 55%
Prove it, then ship it
Build an eval set, secure the platform against injection and data leaks, red-team another team's, then deploy yours to a live URL.
See sessions & lab
Session 1
Evaluate it & secure it
- Building an eval set, and evaluating an agent end to end
- Tracing a run step by step
- Prompt injection, direct and indirect
- PII and data leakage
- Secrets and least privilege for tools
- Human approval on irreversible actions
- Cost-runaway protection
Session 2
Localhost → production
- Development vs production
- Hosting and domains
- Environment variables and secrets
- Open-weight models vs hosted endpoints vs closed APIs
- Rate limits, retries and fallbacks
- Logs and monitoring
Lab · 1.5 hrsRed-team another team's platform, make it leak, overspend or act outside its remit, fix your own, then deploy to a live URL.
Final evaluation · 55%The complete product, assessed end to end: grounded, connected, evaluated, secured and deployed.
The case study
One build, start to finish.
Not weekly assignments. The cohort carries one e-commerce platform from the build day in Goa to the final evaluation, and every session adds the next layer.
Storefront →Database →AI agents →Connections →Security →Deployment
The evaluation
Six things your product must be.
The complete product is assessed end to end against six checks. The bar is shared on Day 1 in Goa.
Evaluation 1 · Week 245%Final evaluation · Week 455%Complete the course45+Below 45One resubmission
01
Grounded
Answers from the store's own data, and can show where it came from.
02
Connected
Reaches real apps and databases over MCP and automated flows.
03
Evaluated
An eval set and traces prove it works, not a demo that happens to.
04
Secured
Survives red-teaming: injection, leakage, secrets and runaway cost.
05
Deployed
Live on a URL anyone can open, with logs and fallbacks in place.
06
Explained
You can defend the stack, the data and the model choice.
What you leave with A product. A certificate.
A product. A certificate.
A shot at an internship.
Course completion certificate
Complete the programme, scoring 45 or above across both evaluations. Issued by Vishlesan i-Hub, IIT Patna's Technology Innovation Hub, in collaboration with Masai.
- Est. 2008 by an Act of Parliament
- Institute of National Importance
- Campus in Bihta, near Patna
You will get access to placement opportunities by Masai after clearing the evaluation.
The founders and investors you meet in Goa, your mentors and your cohort of 50.
in free AI credits for your build. *For selected participants
Who's in the room Founders and investors.
Founders and investors.
In one room, in Goa.
Two ways you get in front of the startup world while you're in Goa.
Cohort founders See what you build
Early-stage founders, building right now
Founders from Masai's Leap Founders Cohort, in Goa to find their problem and build their team. They're in the room to see what you've built.
What you show The live product you built and deployed on Day 2 Your cohort 49 other builders, your pod and your mentor
Investor–Founder Mixer In person in Goa
The room most students never get into
Investors and startup founders, in person. Show them what you built, and hear how they judge a product and the people behind it.
Investors How they look at a product, a market and a team Startup founders How they build, and who they hire
Where you'll stay Your base for 3 days:
Your base for 3 days:
Fairfield by Marriott Goa Calangute.
We've booked the hotel for the whole cohort, so your time goes into learning, building and meeting founders, not logistics.
Booked for the cohort Fairfield by Marriott
View on Google Maps Fairfield by Marriott
Goa Calangute
Gauravaddo, Calangute, North Goa
Your stay, covered
- Twin-sharing rooms for all 3 days
- Every scheduled meal during the residency
- A short walk from Calangute Beach
Twin rooms 
Poolside dining The faculty
Learn AI from people building with it every day.
Practitioners working across AI, cloud, machine learning and product engineering.
Google Aditya Rane
AI & ML Technical Solutions Consultant
Microsoft Deepak Kasera
Senior Engineer, Cloud & AI
Simplismart Shubhendu Shishir
Head of Engineering
Intuit Saksham Arora
Backend Engineering Lead, Identity
PayPal Mohit Uniyal
Senior ML Engineer
Ericsson Amit Singh
Google Authorized Trainer
How to apply ₹99 to register.
₹99 to register.
You pay nothing more unless you're selected.
Applications are reviewed on a rolling basis and seats are limited.
Register for ₹99
Pay ₹99 and tell us about your work or studies and what you want from the residency. This puts you in the review pool.
Applications are reviewed
We look at intent, ambition and whether you're ready to pitch yourself to founders.
Attend your counselling
A conversation with the team about the program, the commitment and the travel to Goa. Pay ₹5,000 at this step.
Confirm your seat
Pay the remaining ₹45,000 + GST to lock your seat, then onboarding: pre-reading, a setup checklist and travel details.
The Fee
What you put in.
One fee covering every session, your stay and meals. You pay nothing beyond ₹99 unless you are selected.
₹50,000 + GST
3 days in Goa + 4 weeks online · Fairfield by Marriott Goa Calangute
₹99 registration₹5,000 at counsellingBalance once you're selected
Included
- Your stay at Fairfield by Marriott Goa Calangute for all 3 days: twin-sharing rooms and all scheduled meals.
- Every Goa session, the build day and the lab, plus 8 live online sessions and 4 labs.
- Templates, resources and learning materials.
- The Investor–Founder Mixer in Goa.
- Course completion certificate from i-Hub, IIT Patna
Not included
- Travel to and from Goa, including airport transfers.
- Any extra stay before or after the programme dates.
- Paid AI tools or subscriptions.
- Laptop and working setup.
Program fee is refundable as per our Refund & Withdrawal Policy
FAQ
Questions worth answering before you apply.
What is the Leap AI Residency, Goa?
3 days residential in Goa, then 4 weeks online, for a cohort of 50. You learn enough technology to build intelligently with AI and build and deploy a live AI product. In Goa you also meet investors and startup founders at the Investor–Founder Mixer.
When does it start?
The Goa residency starts on 20 November 2026 and runs for 3 days. Week 1 online begins the week after you get home.
Is the programme online or offline?
Both. The first 3 days are in person in Goa, and so is the founder pitch. Then 4 weeks online: two live sessions and a 1.5-hour lab each week.
What happens during the 3 days in Goa?
Day 1 you arrive, hear why the programme exists, meet the cohort and hear from an AI and product leader. Day 2 is build day: take apart a real product, then build and deploy your own to a live URL with your pod. Day 3 covers how the web, APIs and databases work, with an evening lab.
What happens in the 4 online weeks?
Week 1: models and prompting. Week 2: turning an idea into a working product with coding agents. Week 3: agents, your own data with RAG, and connecting to other systems over MCP. Week 4: evaluating, securing and deploying. Every session builds on the e-commerce platform you start in Goa.
How much time does the online phase take?
Two live sessions and a 1.5-hour lab every week, for 4 weeks.
How is the work assessed?
Two evaluations: a checkpoint in Week 2 (45%) and a final evaluation in Week 4 (55%). The final product must be grounded, connected, evaluated, secured, deployed and explained. The completion bar is shared in Goa on Day 1. Score 80% or above in the final evaluation and You will get access to Placement opportunities by Masai after clearing the evaluation.
What certificate will I get?
A course completion from Vishlesan i-Hub, IIT Patna, issued by Vishlesan i-Hub, IIT Patna, in collaboration with Masai. You complete the course with a score of 45 or more across the two evaluations; below 45, you get one chance to resubmit.
Who are the founders in the room?
Cohort founders, early-stage founders from Masai's Leap Founders Cohort who are in Goa building right now; they see what you've built. You also meet investors and startup founders at the Investor–Founder Mixer.
What is the Investor–Founder Mixer?
An in-person mixer in Goa with investors and startup founders. You show what you've built and hear how they judge a product and the people behind it.
What's the founder pitch?
In person in Goa, you pitch yourself to the startup founders in the room.
How does the internship opportunity work?
Clear the evaluation at 80% and You will get access to Placement opportunities by Masai after clearing the evaluation.
Will the AI Residency get me a job?
There is no job guarantee. Clear the evaluation at 80% and you get access to placement opportunities by Masai. Everyone leaves with a live AI product they built and can explain.
Who should join the AI Residency?
College students and fresh graduates who use AI but can't yet build with it.
Can college students join?
Yes. College students and fresh graduates are welcome.
Do I need a technical degree?
No technical degree is a pre-requisite. Selection is based more on fit, intent and your ability to commit to the programme.
Do I need to know how to code?
No. You don't need coding experience or a technical degree. You'll learn how software works underneath and how to direct AI coding tools.
I've never pitched before. Can I still join?
Yes. No pitching experience is needed to join.
What will I learn?
How the web, APIs and databases work; what models are and how to control them; building with coding agents; agents, RAG and MCP; and evaluating, securing and deploying an AI product.
Will I actually build something?
Yes. You build and deploy a working product on Day 2 in Goa, then carry one e-commerce platform through the 4 weeks to the final evaluation, live on a real URL.
What kind of AI tools will we use?
AI coding agents such as Cursor, Claude Code and Codex, models on Hugging Face, Git and GitHub, n8n for automation and MCP servers. Specific tools may change as the ecosystem moves.
Do I need my own laptop?
Yes. Bring a personal laptop capable of running the programme tools.
Will there be free time?
The 3 days in Goa are structured and intensive, with sessions filling most of each day. Any free time outside it is yours to use independently and at your own expense.
Where will I stay?
At Fairfield by Marriott Goa Calangute, Gauravaddo, Calangute, North Goa, for all 3 days of the programme.
Is accommodation included?
Yes, for the officially communicated residency dates. Any extra stay before or after the programme is not included.
Can I request a private room?
No. Standard accommodation is twin-sharing.
Are airport transfers included?
No. Travel to and from Goa, including airport transfers, is not included in the programme fee. You get yourself to the hotel; we take it from there.
Will local transport be provided?
Transport will be arranged for scheduled programme activities in the official itinerary. Personal travel outside the itinerary is not included.
Can family or friends join me?
They may travel independently, but companions are not included in programme accommodation, activities, transport or other arrangements.
Do I need to submit medical information before travelling?
Yes. A Medical Fitness Certificate issued by a registered medical practitioner is mandatory before the residency. Any medical consultation, treatment, medication, hospitalisation or related expenses during the residency are borne by the participant and not covered in the programme fee.
Do I need medical insurance?
Yes. You must arrange suitable medical insurance before travelling; it is not covered in the programme fee.
What food options will be available?
Meals follow a pre-defined menu, with vegetarian and non-vegetarian options available. Special meal requirements outside the standard menu must be arranged and managed by the participant at their own expense.
Registrations open Build it.
Build it.
Then show the founders.
50 seats. 3 days in Goa and 4 weeks online. Build a live AI product, meet investors and founders, earn an i-Hub, IIT Patna certificate and your shot at an internship.
₹99 to register. Every application is reviewed personally. Starts 20 Nov 2026.