We teach AI · Not slides about AI

Learn AI to survive any future

Cockroach School teaches AI from absolute basics to training your own models, so no matter how the world changes, you adapt, build, and survive.

6 months
Live-in residency
Madhapur, Hyderabad
Learn by building Train your own models Ship real products
Cockroach School mascot holding a sign
Intelligence under pressure
Built by alumni from
Indian Institute of Technology, Roorkee
Partners & ecosystem
Ovam.ai Pursuit Future Technologies Mom Pharmacy Prov Logic
Programs

One school.
AI first, then the company.

Start here

AI Engineering Residency

6 months · Full-time · Live-in · Beginner & Engineer tracks

From your first line of Python to RAG, agents, deployment and a fine-tuned model. Six shipped projects. This is the core of Cockroach School.

See the curriculum →
Advanced

LLM Training Lab

12 weeks · For engineers · GPU access included

Train language models, not just call them. Data pipelines, pretraining a small model from scratch, supervised fine-tuning, DPO and RLHF, evals, quantisation and serving on GPUs.

Ask about the next batch →
Go-to-market

GTM Lab

8 weeks · LeadWoo account included · Run a real campaign

Learn to find customers, not just build for them. ICP and list building, LinkedIn outreach, cold email and deliverability, WhatsApp follow-ups, Meta lead ads, CRM pipeline, and getting found by AI search. You run it all on LeadWoo for a real business, and keep the account through the course.

Ask about the next batch →
After you ship

Cockroach Founders

12 weeks · For residency graduates and working builders

The entrepreneurship program. Turn the AI product you shipped into a company: first customers, pricing, sales, incorporation, fundraising, demo day.

Ask about the next batch →
Optional

Degree pathway

Alongside any program

We don't award degrees. We help you enrol in a UGC-recognised online degree from a university that teaches and awards it directly, and we schedule the residency around it.

The deal

Ship a live AI product in your first 30 days.
Or you don't move to month two.

Day 1

Laptop open. Python, VS Code, your first API call to an LLM. No theory week.

Day 30

Your first AI product is deployed on a real URL with real users giving you real feedback.

Day 180

Six shipped AI projects including a model you trained yourself, a production-grade portfolio, and the AI engineering interviews to show for it.

Why Cockroach

AI first.
Then the company.

Everything starts with building AI: Python, LLM APIs, RAG, agents, deployment, then training your own models. Only once you can ship do we teach you how to find customers, sell it, and build a company around it.

01

Learn by building

Every module ends with something deployed, not a quiz. You learn RAG by shipping a RAG app, agents by shipping an agent.

02

Live where you build

Beds on campus in Madhapur, minutes from Hitech City. Your desk, your bunk, and your cohort under one roof.

03

The real AI stack

Python, FastAPI, OpenAI / Claude / Gemini APIs, embeddings, vector databases, PyTorch and Hugging Face, Docker and cloud deployment. The tools teams actually hire for.

04

Taught by people who ship AI

Instructors who build AI products for a living, plus sessions with founders and operators who use AI every day. Ask anything, get real answers.

05

Hitech City internships

Work on live problems inside startups next door while you study, so your first job is not your first job.

06

Keep your degree on track

The residency fits around a UGC-recognised online degree, so you don't choose between a qualification and real skills.

Curriculum

Choose your AI survival path

A battle-tested roadmap to become an AI engineer, whether you are starting from scratch or upgrading from software engineering.

Who it's forBeginners or non-programmers who want to become AI engineers. Students, career switchers, non-tech learners.
OutcomeJob-ready AI engineer capable of building RAG apps, simple agents and fine-tuning small models, with a capstone product live on the internet.

Module 0Foundations

  • Programming mindset
  • Python & VS Code setup
  • Mini projects

Module 1Python essentials

  • Python basics & OOP
  • File & data handling
  • Python for AI, not generic Python

Module 2Core LLM & RAG skills

  • LLM fundamentals
  • Prompt engineering
  • RAG systems
  • Evaluation basics

Module 3AI product foundations

  • FastAPI basics
  • AI-powered APIs
  • Deployment fundamentals
  • Logging & monitoring

Module 4Intro to AI agents

  • Agent fundamentals
  • Tool usage
  • Simple agent architectures

Module 5Train your first model

  • How models learn: data, loss, gradients
  • PyTorch basics & Hugging Face
  • Fine-tune a small open model with LoRA
  • Evaluate it, then serve it

Module 6Career, portfolio & capstone

  • Industry roles explained
  • Capstone AI project
  • Resume & LinkedIn
  • Interview readiness
Who it's forFull-stack, backend and software engineers shifting into AI engineering. Developers already in tech who want production-level AI skills.
OutcomeBuild production-ready RAG systems and AI agents, train and fine-tune your own models, deploy real AI products, and interview at a senior level.

Module 0Modern AI stack

  • APIs, embeddings, vector DBs, RAG, agents
  • Python + FastAPI project setup

Module 1Python for AI engineering

  • Python speed-run for engineers
  • OOP, packaging, utilities
  • Async + HTTP basics
  • Minimal NumPy / Pandas
  • Logging & tooling

Module 2Core AI engineering

  • LLM fundamentals & prompt design
  • OpenAI / Claude / Gemini APIs
  • Embeddings & vector databases
  • RAG pipelines & evaluation

Module 3AI product engineering & deployment

  • CI/CD & testing AI systems
  • Observability & logging
  • Cost optimisation
  • Docker & cloud deployment

Module 4AI agents (advanced)

  • Tool usage & orchestration
  • LangChain, LangGraph concepts
  • From-scratch agent systems
  • Safety & reliability

Module 5Model training & fine-tuning

  • PyTorch, transformers from the inside
  • Datasets, labelling, synthetic data
  • LoRA / QLoRA fine-tuning on a GPU
  • Evals, quantisation, serving with vLLM
  • When to fine-tune vs RAG vs prompt

Module 6Portfolio & career systems

  • Portfolio standards
  • Senior-level project design
  • AI interview preparation
  • Freelancing & career strategy
What you will ship

Six things on the internet with your name on them

Project 1

AI Support Copilot

Project 2

Document Q&A System

Project 3

AI API Backend

Project 4

Autonomous AI Agent

Project 5

Your own fine-tuned model

Capstone

Production-ready AI Portfolio

Tools you will master

The stack, not the hype

PythonFastAPIOpenAI APIClaude APIGemini APIPrompt engineeringEmbeddingsVector databasesRAG pipelinesPyTorchHugging FaceLoRA / QLoRAGPU trainingDatasets & labellingvLLMLangChainLangGraphAI agentsEvaluationDockerCloud deploymentCI/CDObservabilityGit & VS Code
Campus

A floor built for building.
Beds upstairs.

Cockroach School classroom floor in Madhapur with a cohort in session

Classroom floor

Open-plan desks, whiteboards everywhere, fast internet. Sessions run here, so does your late-night debugging.

Beds on campus

Dormitory beds a staircase away from your desk. Rent is separate from the program fee, and you can opt out if you live in Hyderabad.

Hitech City next door

5th Floor, RR Building, 100 Feet Main Road, Madhapur. Walkable to the startups and offices you'll intern at.

Cockroach School team at the campus
Learn from doers

Cockroach on the road

Learn online. Connect offline. The Cockroach community turns up where the people using AI at scale are, asks hard questions, and brings the lessons back to campus.

Cockroach community member with Jayakumar P
Jayakumar PChief Executive Officer, Apollo Pharmacy
Cockroach community member with Piyush Shaw
Piyush ShawCo-founder, InMobi · President & COO, Glance
Cockroach community member with Ronnie Screwvala
Ronnie ScrewvalaChairperson & Co-founder, upGrad
Cockroach community member with Pavan Guntupalli
Pavan GuntupalliCo-founder, Rapido
Cockroach School at Global Fintech Fest 2025 Cockroach School at Global Fintech Fest 2025
On the road

Global Fintech Fest 2025, Mumbai

The cohort doesn't wait for industry to visit. We go to it: conferences, demo days and founder meetups, with notes shared back to everyone on campus.

Simran Kour Sidhu
Founder spotlight
“Cockroach School throws you into chaos. Real problems, real pressure, real learning. Then it teaches you how to survive and win.”
Simran Kour Sidhu
Co-founder, Ovam.ai · IIT Roorkee
Admissions

How to get in

Beginner trackClass 12 or above, any stream. No coding experience needed. You need a laptop and the stubbornness to finish.
Engineer trackWorking developers, or graduates who can already build and deploy a basic web app.
RoundDeadlineCohort starts
Early bird——
RegularUntil seats fill—

Program fee and bed pricing are shared on the call. Scholarships for applicants who have already shipped something, anything, on their own.

Tell us what you want to learn

Fill the short form below. Name, number, and one honest paragraph.

A 20-minute call

We talk about where you are, which track fits, and whether you'll actually enjoy six months of this.

The 48-hour build

No aptitude test. You get a small brief and two days. Beginners get a no-code version. We look at how you think, not what you know.

Offer and move-in

Seat confirmed, bed allotted if you want one, and a pre-work pack so Day 1 starts fast.

FAQ

Questions people ask

I have never written code. Is this for me?

Yes. The beginner track starts at Module 0 with the programming mindset and setup. Half the cohort usually starts from zero.

Is it really full-time and residential?

Full-time, yes. Residential is optional. Beds on campus are priced separately from the program, and Hyderabad locals can commute.

Do you teach model training or only APIs?

Both. You start by building on OpenAI, Claude and Gemini APIs because that is how most products ship. Then you go under the hood: PyTorch, Hugging Face, fine-tuning open models with LoRA on a GPU, evaluating and serving them. You leave knowing when to train and when not to.

What do I actually leave with?

Six deployed projects with live URLs, including a model you fine-tuned yourself, a portfolio built to a standard hiring managers recognise, and interview preparation for AI engineering roles.

Does this replace a degree?

No, and it doesn't have to. The schedule is designed to fit around a UGC-recognised online degree if you want to keep one running.

Is this an entrepreneurship course?

Not first. The AI Engineering Residency and the LLM Training Lab come first. Then GTM Lab teaches you to find customers with LeadWoo, and Cockroach Founders is the entrepreneurship program for people who want to build a company around what they shipped.

Can I get a degree?

Cockroach School doesn't award degrees. Like other new-age schools, we help you pair the residency with a UGC-recognised online degree from a university that teaches and awards it directly, so you leave with both.

Why cockroaches?

Because they survived everything the planet threw at them. Dinosaurs didn't. Adapt, build, survive.

Chat with us

Survival-first starts here

Tell us what you want to learn. A human from the team replies, usually the same day.