Case study
AI Integration Case Studies
Break down how a real company put AI into its product or operations, and what it changed.
AI First Learning
Learn AI by applying it to real business problems, building products, and shipping solutions.
The AI curriculum, in numbers.
- 150hrs
Of AI learning and building
- 25+
AI tools students get fluent in
- 6
Terms, from no-code to fine-tuning
- 50+
AI products launched by the Mesa students in year one
The Mesa approach to learning AI.
Every Student Builds with AI
Every student builds with AI, with projects tailored to their baseline and target role.
Business + Technical Mentorship
Learn from both business and technical mentors at every AI workshop and hackathon.
Showcase Your AI Builds
Present projects to external founders and operators every term. Top builds get featured publicly.
Feedback in 48 Hours
Get actionable feedback on every submission within 48 hours.
Six terms, one path to your own AI product.
- Term 1
Foundations of AI & No-Code Product Building
- Term 2
Voice, Multimodal, and Agentic AI Products
- Term 3
Vertical AI Products + Automation + Vibe Coding
- Term 4
Fine-Tuning & Open Source AI Productization
- Term 5
Building & Scaling AI Agents and Systems
- Term 6
Thesis: Solve a Real Problem with AI
- The goal
Launch publicly and reach 100 paying customers
The thesis product goes live on Product Hunt and is pitched to investors.
AI Projects & Challenges
Every term ends in builds that are tested on real companies, real users and real revenue, plus a hands-on AI Building Test.
2–4 weeks
Live Projects with Companies
Take a real brief from a partner company and hand back a working AI build.
Hackathon
Sectoral-Functional Hackathons
Build a working AI solution for one sector or function. A product, not a deck.
Challenge
GTM Challenges
Take an AI product to market: position it, price it and find its first users.
Challenge
What AI Features Would You Add?
Pick a product you use every day, then design and prototype the AI features it is missing.
Challenge
User Challenges
Put your build in front of real users, watch what they do and ship the fixes.
Challenge
Monetising Challenges
Turn a build into revenue, from pricing and a paywall to the first paying customer.
AI Masterclasses
Every term, founders and operators unpack real AI products, sharing how they're built, deployed, and scaled.
Behind the Scenes of an AI Product
See how a live AI product really works, and what broke.
AI Product Architecture Breakdown
Tear down a real AI product's stack and moat.
Voice AI Sprint
Build a working voice agent, from speech-to-text to speech.
GTM for AI Products
Position and sell an AI product the AI-native way.
Sales for AI Products
Handle AI objections and run a POC that closes.
AIxSector Deep Dives
Find where AI is creating roles in finance, health, legal, ops and more.
Marquee Mentors
Operators who build with AI every day run the workshops, sit in on hackathons and judge the AI Showcase.
Explore workshopsNaveen Kumar Bhansali
Founder, BlitzAI
AI expertise
AI & Analytics Consultant with experience in applied AI and analytics; associated with IIM Bangalore.
Blessin George Varghese
Seasoned AI Practitioner
AI expertise
Works with LLMs, conversational AI, and AI security, with experience exploring frontier AI technologies.
Bhaskarjit Sarmah
Head of AI Research, Domyn
AI expertise
Leads financial AI research and mentors learners in Generative AI, LLMs, and machine learning.
Vishal Gupta
AI & GenAI Product Leader, Haleon
AI expertise
Builds and scales AI-powered products, from early-stage innovation to enterprise-level deployment.
Harkirat Singh
Product Manager & GTM Lead, Zivy
AI expertise
Builds technology products and supports AI founders through buildathons, product demos, and builder communities.
Dhruv Dhingra
Business & Growth Operator, Lokal
AI expertise
Teaches practical applications of AI in business, growth, and operational decision-making.
Tushar Nain
Product Lead, Grapevine
AI expertise
Built AI-powered hiring products, including a job discovery agent that reached 100K users.
Mitul Aggarwal
Founder & Chief Architect, 302Labs
AI expertise
Builds AI-powered business automation systems and has delivered AI training at Mesa School of Business and IIM Bangalore's NSRCEL.
Divij Bajaj
Data & Applied Scientist, Microsoft
AI expertise
Builds production-scale AI/ML solutions and award-winning AI agents, including an AI education copilot.
Sreeram Sridhar
AI Product Manager, Swiggy
AI expertise
Builds enterprise GenAI products and agentic AI workflows for large-scale business operations.
25+ AI tools. The stack real startups run on.
Students get fluent in the tools AI-native teams use every day, from no-code builders to agent frameworks. The stack is reviewed every cohort.
Claude
ChatGPT
Perplexity
Gemini
Cursor
LLovable
Replit
v0
Vercel
GitHub
Supabase
Hugging Face
Ollama
LangChain
n8n
Make
Zapier
RRelevance AI
ElevenLabs
VVapi
Gmail
Google Sheets
Slack
Notion
HubSpot
PostHog
PPinecone
Resend
Razorpay
Figma
Product Hunt
BYOP: Build your own product
Build a full-stack AI product in 8 weeks.
For those looking to move closer to technology and into AI-first roles. Build and deploy a real AI product, test it with real users, measure its impact and develop a business model to monetise it, creating proof of work for your next role.
Judged at the AI Showcase by outside founders and operators
- Build
- Deploy
- Real users
- Monetise
- Step 01
Scope
Pick a problem you have seen yourself: at work, in an internship, or in your BSL project.
- Step 02
Build
Design the workflow, add a simple front end and a database.
- Step 03
Ship
Deploy it as a live link anyone can open.
- Step 04
Get users
Find real users. Report what worked, what broke and how many used it.
- Step 05
Monetise
Work out how it could earn, and present the business model.
From the BYOP sprints
Pitch day
BYOP pitch day
Cash prizes for the top AI products of the sprint.
Year one
50+ AI products launched
By Mesa students in their first year.
₹10L/month
Up from ₹1L a month, between December and January
AI Job Hunter
Built with Claude Code by Ashutosh Lath.
- Claude Code
1,300
Users in its first 7 days
Snowball
An AI retirement planner taken from MVP to pitch-ready in under 6 weeks, by Pratik Rajpurohit and Vardhraj Songara.
- Replit
- LLovable
8+
Companies sold to
Aatreus
AI-native SaaS that turns messy sales conversations into clear next steps for Indian B2B teams. Built by Dhroov Joshi.
Student showcase
From idea to AI product: what Mesa students have built
01Sales
An AI sales co-pilot
Listens to calls and surfaces real-time answers from your company documents.
02Finance
A finance approval dashboard
Streamlines fund requests with one-click approvals or clarifications.
03Operations
Automatic reimbursement emails
Extracts bill details and drafts ready-to-send reimbursement emails.
04Marketing
A generative engine optimisation dashboard
Shows brands how visible they are inside AI search results.
05Supply chain
A supply-chain optimiser
Maps customer requirements to the fewest suppliers for maximum fulfilment.
06Customer
A sentiment and review analyser
Summarises customer feedback and flags product issues.
07Operations
An order-management dashboard
Auto-organises customer orders, preferences and payments.
08Retail
A digital order-management dashboard
Auto-organises customer orders, preferences and payments in one place.
09Personal finance
An AI retirement planner
Builds personalised SIP and long-term wealth plans instantly.
10Personal finance
A credit card recommender
Picks the best card based on spending habits, rewards and eligibility.
11Content
An AI content engine
Builds your unique writing style guide and recreates posts in your tone.
12Careers
An AI job-application platform
Tailors resumes, scrapes listings and sends personalised cold emails.
13Careers
An AI interview trainer
Helps people master communication and self-presentation.
14Careers
A mock interview simulator
Analyses your LinkedIn and simulates real interviewer personas.
15Health
A fitness app
Finds people with your body type and reverse-engineers a plan for you.
16Education
An adaptive SAT prep engine
Builds a personalised maths roadmap with diagnostics and smart mock tests.
17Education
An AI essay facilitator
Helps IBDP students prepare, correct essays and track progress.
18Wellbeing
A Gita chatbot
Gives verse-backed, contextual guidance tailored to the user.
Inside the build sessions
What students and parents ask.
Do I need to know how to code?
No. Term 1 starts from zero, with prompt engineering, reading API docs and no-code app deployment. Vibe coding comes in Term 3, and fine-tuning and multi-agent systems follow from Term 4.
Is this theory or building?
Building. Every course ends in something that works, and every build is deployed as a live link. Concepts are taught so students can build better, not to be memorised.
Won't the tools be out of date in four years?
Some will. That's why we review the stack every cohort and teach the thinking underneath it: how to break a problem down, describe it clearly, verify outputs and design systems that don't break. Those skills transfer to whatever tool comes next.
How much of the programme is AI?
The AI curriculum is 150 hours across six terms, running from no-code tools to fine-tuning, and students become fluent in 25+ AI tools along the way.
Who judges the work?
Practising operators and outside founders. They judge the AI Showcase each term and give the kind of feedback students would get on the job.
Find a problem. Build the workflow. Show the impact.
Six terms of AI, a product of your own, and Demo Day to pitch what you built.