CloudxLab × IIT Roorkee Certificate Program
Flagship Cohort Program

The Inventor's Journey
From First Principles to AI

Stop memorising formulas. Start inventing.
Build your own number systems, algorithms, ML models, and neural networks - then understand why they work. Jointly certified by IIT Roorkee and taught by IIT alumni and IIT Professors.

📅 10 Months
🕐 Weekends · 2 × 2–3 hr sessions
🏛️ IIT Roorkee Campus Immersion
📜 IIT Roorkee Certificate
🧪 30+ Projects
☁️ Cloud-Based Lab
🎓 10th Grade Onwards
✗ 30 May · 8–10 pm IST · Sold Out ● 30 Jun · 8–10 pm IST · Enrolling Now ● 30 Sep · 8–10 am IST · Open
✦ Apply for Jun 2026 Batch Explore Curriculum
9
Deep Modules
30+
Projects
IIT
Roorkee Cert.
10mo
Weekend Pace
6+
Career Paths
🏛️ IIT Roorkee Certificate
🎯 First-Principle Learning
🤖 AI Agents & LLMs Included
🧪 30+ Hands-on Projects
📅 Weekend Schedule
✈️ Campus Immersion
☁️ Cloud Lab Included
🐍 PyQuest for Beginners
Our Promise

This is not a course.
It is a transformation.

Most programs hand you recipes. We hand you problems and step back. By the time you finish, you won't just know how algorithms work - you will know why they exist, because you invented your own versions of them first.

"Their education trained them to receive knowledge, not generate it. When the answer key disappeared, they froze. That's the gap this course closes."
📅
Duration
10 Months
Weekend pace without pausing your degree or job
🕐
Schedule
Weekends
2 sessions/week · 2–3 hrs each
📜
Certificate
IIT Roorkee
Recognised across industry
🏛️
Campus
IIT Roorkee
On-campus immersion included
🧪
Projects
30+
One per topic, you invent your own
🎓
Eligibility
10th Grade+
No prior coding needed
☁️
Lab
Cloud-Based
No setup, code from any browser
🐍
Beginners
PyQuest
pyquest.io, Python-ready before day one
Your Instructors

Learn from practitioners & researchers

IIT alumni, IIT professors, a NASA research scientist, Yale PhDs, and startup founders - faculty who have built real systems and know how to teach you to build yours.

Collaborative learning environment
Teaching that transforms, not just informs.
Sandeep Giri

Sandeep Giri

Lead Instructor · Course Creator

Founder of Terno AI. AI/ML Researcher with industry experience at Amazon, InMobi, and D.E. Shaw. Graduate of IIT Roorkee.

IIT Roorkee AlumniAmazonD.E. ShawInMobiFounder, Terno AI
LinkedIn
Sanjeev Manhas

Sanjeev Manhas

Professor, ECE Department · IIT Roorkee

Faculty at the ECE Department, IIT Roorkee. Brings academic rigour and first-principle derivations to the Machine Learning module.

IIT Roorkee FacultyECE Department
LinkedIn
Abhinav Singh

Abhinav Singh

Co-Founder · CloudxLab

Co-Founder of CloudxLab. Previously at BYJU'S, scaling educational technology for millions of learners.

CloudxLab Co-FounderBYJU'S
LinkedIn
Dr. M.L. Virdi

Dr. M.L. Virdi

Senior Research Scientist · NASA

Senior Research Scientist at NASA. Brings world-class experience in applied science, AI, and space exploration.

NASAResearch ScientistApplied AI
LinkedIn
Praveen Pavithran

Praveen Pavithran

Co-Founder · Yatis

Co-Founder of Yatis. Previously at YourCabs and Cypress Semiconductor. Bridges academic knowledge and real engineering demands.

YatisYourCabsCypress Semiconductor
LinkedIn
Jatin Shah

Jatin Shah

Yale CS Ph.D. · IIT Bombay · Ex-LinkedIn & Yahoo

CS Ph.D. from Yale University, undergraduate from IIT Bombay. Industry experience at LinkedIn and Yahoo.

Yale CS Ph.D.IIT BombayLinkedInYahoo
LinkedIn
Learning Philosophy

What is the Invent-to-Learn Method?

Struggling with a problem you designed yourself creates understanding no lecture ever can. Every principle below was shaped by one conviction: the most valuable learning environment is one where no answer is handed to you.

Student learning by doing

The best way to understand something is to build it yourself.

Problems first. Theory second. Invention always.

🔢

Start with your own number system

Day one, you design your own base and do arithmetic in it. This single exercise rewires how you think about abstraction, forever.

🧪

Problems before explanations

Simple problem statements gradually grow in complexity. Theory is introduced only once you've felt the need for it, making it stick.

⚙️

Invent your own algorithms

You'll build your own decision tree, gradient descent, KNN, and neural network - before ever reading a textbook definition.

🙋

Questions are celebrated

Curiosity is the curriculum. There are no dumb questions. You'll learn to enjoy, even celebrate questions from classmates.

🎤

Defend Your Invention

You don't just build - you present and defend your approach to the class. Teaching what you built is the deepest form of understanding.

🎯

Minimal Theory, Maximum Insight

We teach only the key ideas. The obvious parts are deliberately left for you to figure out - because that act is precisely where deep understanding is forged.

🎉

Nudge silently. Celebrate loudly.

When you're stuck, a quiet reframe protects your confidence. When you solve something, it is acknowledged loudly. That changes everything.

🪞

Blame no one, own your gap

Looking inward when something doesn't make sense - instead of blaming the explanation or your ability - is the most important meta-skill a learner can develop.

The Instructor's Beliefs

How Sandeep teaches, in his own words

Not teaching policies. Convictions shaped over 12 years of teaching and 400,000 learners. Read more →

"If you don't understand, that's on me, not on you."

When a student can't grasp a concept, the first question is: what did I get wrong? The explanation, the sequence, the example, something in the teaching failed. This constraint forces every explanation to keep improving until it lands.

"The goal is never to make the problem feel hard. The goal is to make it feel solvable."

When you're stuck, Sandeep doesn't repeat the same explanation louder. He reframes - simpler words, a smaller piece, a different angle - until the problem is within reach. Then he steps back and lets you solve it.

"Understand them enough to find their strength, then hold it up where they can see it."

Real mentorship begins with genuine recognition of what is already there. A student who knows their own strengths becomes a different kind of learner: curious instead of anxious, persistent instead of avoidant.

What You'll Build

Full Curriculum: 9 Modules

Every module ends with projects you invented, not copied.

Already studying Python or DSA in college? We cover the same topics, but go 10× deeper by making you invent your own solutions before ever seeing the textbook answer.
★

Linux, SQL & Git

Self-paced · Complete at your own schedule alongside the main program

Self-Paced
Linux fundamentals: command line, file system, shell scripting
SQL: querying databases, joins, aggregations, and real-world data analysis
Git & version control: commit, branch, merge, and manage your 30+ projects from day one
0

Designing Your Own Number System

Prologue · Sets the tone for the entire journey

Prologue
Choose your own base; write and count up to three digits in it
Perform addition, subtraction, multiplication, and division in your chosen base
Convert numbers between your base and base-10, and discover why base-2 powers every computer
First taste: you are an inventor, not a consumer of knowledge
1

Python by Invention

Python · Maths · DSA · Basic ML · LLMs · AI Agents

Foundation
Math Primer: matrices, logarithms, and basic probability alongside Python, not before it
Python fundamentals through progressively harder maths problems
OOP: build classes, objects, and reusable abstractions
Big O & Complexity Analysis: reason about why one algorithm beats another
Data structures you build yourself: BSTs, queues, trees
Algorithms from scratch: binary search, BFS, DFS, merge sort, quicksort
Dynamic Programming: solve complex problems by breaking into overlapping subproblems
Build and train your own KNN, K-Means, Naïve Bayes, and Decision Tree
Interact with LLMs; Sentiment Analysis with LLM Embeddings
Semantic search and building your first AI Agent
RAG (Retrieval-Augmented Generation): build your own document Q&A system
2

Scientific Python & Data Wrangling

NumPy · Pandas · Matplotlib · Projects

Tools
NumPy, Pandas, Matplotlib: learned by building, not reading docs
Data cleaning and handling messy real-world datasets
Feature engineering: create, transform, and select the right inputs for your models
Build a recommendation engine from first principles
Implement your own Linear Regression with your own Gradient Descent
Interactive data visualisation - tell stories with your data using Plotly
3

End to End Machine Learning

Delivered by IIT Professors · Invent your own algorithm

IIT Faculty
Complete end-to-end ML project: data → model → evaluation → deployment
Linear & Logistic Regression: derive and implement
Decision Trees and Ensemble Learning (Bagging, Boosting, Random Forest)
Model evaluation: Precision, Recall, F1, ROC-AUC
MLOps: deploy models as REST APIs with Flask, containerise with Docker
Compete on Kaggle with real messy datasets
Everyone invents their own variant of a classical algorithm as capstone
4

Foundations & ANNs in Deep Learning

Build an ANN from scratch in NumPy

Deep Learning
Introduction to Artificial Neural Networks: implement one from scratch
Backpropagation, vanishing gradients, optimisers
Regularisation: Dropout and Batch Normalisation
Transfer learning and fine-tuning pre-trained models
5

Sequence Models & Attention

RNN · LSTM · Attention · Autoencoders

Deep Learning
RNNs and LSTMs: time series forecasting on real-world data
Encoder-decoder architecture: sequence-to-sequence models
Attention mechanisms: self-attention, cross-attention, the foundation of Transformers
Autoencoders for representation learning and generation
6

Transformers & Large Language Models

Transformer Architecture · BERT · GPT · Fine-tuning

Deep Learning
Build the Transformer architecture from scratch: encoder, decoder, multi-head attention
BERT and GPT: how pre-training and fine-tuning work under the hood
Fine-tune a pre-trained model on your own dataset
7

Convolutional Neural Networks

Derive convolutions · Design your own architecture

Deep Learning
Convolutions: derive from first principles, implement, and experiment
CNNs for image classification and object detection
Transfer learning: VGG, ResNet, EfficientNet on your own problems
GANs: build a generator and discriminator from scratch
Design your own CNN architecture and defend every decision
8

Reinforcement Learning

Design your own environment · Train your own agent

Advanced
Markov Decision Processes, policies, and reward shaping
Q-Learning and Deep Q-Networks from scratch
Design your own environment and train an agent on it
Capstone: end-to-end project you define, build, deploy, and defend
30+ Projects

Build things that actually work

Every module includes hands-on projects. Here's a taste of what you'll invent.

🔢

Custom Number System

Arithmetic in a base you designed

🌳

Decision Tree from Scratch

Build, train and visualise it in Python

🤖

AI Agent

Your own agent using LLM APIs

🔍

Semantic Search Engine

LLM embeddings + similarity search

🎬

Recommendation Engine

Built on NumPy from first principles

📉

Gradient Descent

Your own implementation of linear regression training

🧠

Neural Network in NumPy

Forward pass, backprop, training loop

🎮

RL Agent

Custom environment + trained DQN agent

🖼️

Image Manipulation

CNN-powered feature extraction

⚗️

Your Own ML Algorithm

Invent a variant during the IIT ML module

💬

Sentiment Analyser

LLM embeddings + classifier you built

🔀

Sort & Search Visualiser

Animate the algorithms you coded

📄

RAG Document Q&A

Retrieval-augmented system answering questions from your own documents

⚡

Transformer from Scratch

Encoder, decoder, multi-head attention: built and trained in PyTorch

🐳

Dockerised ML API

Wrap your trained model in a Flask REST API and ship it in Docker

🏆

Kaggle Competition

Submit to a real competition, climb the leaderboard, build your public profile

+ 14 more projects across all modules

Key Highlights

What sets this program apart

Learning Experience

🔬

Learn by Inventing

Every concept is discovered through structured problem-solving. You build before you read, always.

🧠

First-Principle Thinking

The real outcome is the ability to face any hard problem and know you can figure it out from scratch.

🎯

Minimal Theory, Maximum Insight

We teach only the key ideas. The obvious parts are deliberately left for you to work out.

🤖

AI & LLM Native

Build AI Agents, run semantic search, perform sentiment analysis with LLM embeddings from Module 1.

Platform & Tools

☁️

Cloud-Based Online Lab

No setup. No installations. Every exercise runs in a cloud-based lab accessible from any browser.

⚡

Self-Paced Automated Assessments

Practice on your own schedule with automated hands-on assessments giving instant feedback.

🐍

PyQuest for Absolute Beginners

Never written a line of Python? Start with pyquest.io - our dedicated app that takes you from zero to Python-ready before the course begins.

Credentials & Career

🏛️

IIT Roorkee Certificate & Campus Immersion

Earn a recognised certificate and experience on-campus immersion at IIT Roorkee.

👨‍🏫

IIT Faculty for ML

The ML module is taught by Sanjeev Manhas (Professor, ECE, IIT Roorkee) - academic rigour meets real-world depth.

💼

Interview-Ready

Crack interviews for SDE, Data Scientist, ML Engineer, AI Engineer, ML Researcher, and ML Scientist roles.

IIT Roorkee Campus

What is the IIT Roorkee Campus Immersion?

As part of this program, you'll experience a campus immersion at IIT Roorkee - India's oldest technical institution. Learn alongside peers, interact with faculty, and graduate with a certificate that carries real weight.

🏛️ IIT Roorkee Campus Visit 📜 Official Certificate 👨‍🏫 Face Time with IIT Professors 🤝 Network with Cohort Peers 🎓 Graduation Ceremony

Campus fees ~₹2,000/day charged separately. Travel and accommodation at your own cost.

🌅
Day 1
Campus orientation, lab sessions with IIT faculty, deep-dive workshops
🎓
Day 2
Capstone project presentations, peer demos, graduation ceremony and certificate distribution
Career Outcomes

Which jobs can you get after completing this course?

This program builds the depth to compete for the most sought-after technical roles.

💻 Software Development Engineer
Build scalable systems; strong DSA and systems thinking required
📊 Data Scientist
Blend statistical rigour with ML to drive decisions
⚙️ Machine Learning Engineer
Train, optimise, and deploy ML models at production scale
🤖 AI Engineer
Build AI-powered products using LLMs, agents, RAG, and fine-tuning
🔬 ML Scientist
Research and develop novel algorithms; bridge academia and applied AI
🎓 ML Researcher
Push the boundaries of what AI can do; publish, invent, and advance the field
Who Should Join

Who is eligible for The Inventor's Journey?

🎒

Students

High-school to college. Build a deep technical foundation and enter placement season with a portfolio and an IIT Roorkee certificate your classmates won't have.

💻

Working Professionals

Pivot into AI/ML or sharpen your technical edge. Weekend schedule fits your life.

🔄

Career Changers

From any background. If you can think, you can learn to invent here.

🧩

Curious Minds

You just need the itch to understand how things work - the rest follows.

Classroom Culture

Everyone belongs here

You will find people from various walks of life - 1st-year students sitting alongside 10-year industry veterans. That mix is intentional. Leave mental blocks around age, profession, and experience at the door. Be fearless while asking questions. No question is too basic. No curiosity is too ambitious.

All Genders Welcome All Age Groups All Professions No Prior Coding Needed 10th Grade Minimum Questions Celebrated

What is the certificate like?

Certification Authority

IIT Roorkee Logo

E&ICT Academy, IIT Roorkee

Sponsored by the Ministry of Electronics and IT, Government of India.

CloudxLab

A deep-tech learning platform empowering engineers with real-world AI and Data Science experience.

Get This Certificate

Campus Immersion Program

Experience IIT Roorkee beyond the classroom: connect, present, and celebrate your journey on campus.

What You Experience

1:1 interaction with Professors and Industry Experts
Certificate award ceremony at IIT Roorkee
Network with peers from across the country
Showcase your capstone project to faculty
A once-in-a-lifetime campus experience

Invest in Your Future

Fee Structure

Sold Out 30 May 2026 8-10 pm IST · Sat/Sun
Enrolling Now 30 Jun 2026 8-10 pm IST · Sat/Sun
Open 30 Sep 2026 8-10 am IST · Sat/Sun
Early bird offer ends in --
₹1,99,999 $3,999 Save ₹1,00,000 Save $1,850
₹99,999
$2,149
UG Certificate with IIT Roorkee · All-inclusive · No hidden charges
Base program fee₹1,99,999$3,999
Scholarship (up to)- ₹25,000- $250
Student discount- ₹75,000- $600
Effective program fee₹99,999$2,149
Pay in full and save an extra 5% One-time payment: ₹94,999$2,041 - no EMI, no fees.
Eligibility: UG graduates (B.Tech, BCA, BSc, or equivalent). Receive an official Under-Graduate Certificate from IIT Roorkee.
Apply Now - Jun 2026 Batch

EMI Options (No-Cost)

Pay ₹9,999$299 upfront. Remaining ₹90,000$1,850 in easy no-cost EMIs.

Admission fee₹9,999$299
Monthly EMI₹8,000/mo$185/mo
Processing feeNone
Total payable₹99,999$2,149
Download Brochure
₹99,999 $1,999 Save ₹20,000 Save $400
₹79,999
$1,599
CloudxLab Certificate · All-inclusive · No hidden charges
Admission fee₹9,999$199
Program fee₹69,000$1,400
Effective program fee₹79,999$1,599
Pay in full and save an extra 5% One-time payment: ₹75,999$1,519 - no EMI, no fees.
Eligibility: Open to anyone who has completed 10th grade. Same curriculum, same campus, same IIT Roorkee certificate.
Apply Now - Jun 2026 Batch

EMI Options (No-Cost)

Pay ₹9,999$299 upfront. Remaining in easy no-cost EMIs.

Admission fee₹9,999$299
Monthly EMI₹7,000/mo$145/mo
Processing feeNone
Total payable₹79,999$1,599
Download Brochure

What's Included in Both Tracks

📚 All 9 modules + self-paced SQL/Linux/Git
☁️ Cloud lab access (10 months)
🎙️ Live sessions, recorded replays
🏛️ IIT Roorkee campus immersion
📜 Official IIT Roorkee certificate
🐍 PyQuest beginner prep access
💼 Placement Eligibility Test (PET)
📱 WhatsApp community (lifetime)
🤖 AI-generated session summaries
🏆 Kaggle competition guidance

Ready to become an inventor?

10 months. Weekends. IIT Roorkee certificate. 30+ projects. A changed perspective.
Start before your peers do - your college classmates will still be memorising when you're inventing.

Apply for Jun 2026 Batch
FAQs

Do I need any prior programming experience?

No. The course starts from absolute zero, including designing your own number system. If you completed 10th grade, you have everything you need.

What is the schedule like?

Classes run every Saturday and Sunday, 2 hours per session. The June batch runs 8–10 pm IST, ideal for working professionals. The September batch runs 8–10 am IST. The 8-month weekend pace fits comfortably around a full-time job or studies.

What does the IIT Roorkee Certificate mean?

On successful completion, you receive an official certificate from IIT Roorkee - one of India's premier technical institutions - along with an on-campus immersion experience.

Who teaches this course?

Led by Sandeep Giri (IIT Roorkee alumnus, founder of Terno AI, ex-Amazon, InMobi, D.E. Shaw). Faculty also includes Sanjeev Manhas (Professor, ECE, IIT Roorkee), Dr. M.L. Virdi (NASA), Abhinav Singh (Co-Founder, CloudxLab), Praveen Pavithran (Co-Founder, Yatis), and Jatin Shah (Yale CS Ph.D., IIT Bombay, ex-LinkedIn, ex-Yahoo).

What does "invent your own algorithm" mean in practice?

You receive a problem statement and constraints. Before seeing any standard solution, you design your own approach, code it, and defend it - then compare with established methods.

How many projects will I complete?

More than 30 projects across all 9 modules. Every project is something you invented, not copied.

Are there EMI options?

Yes. EMI is available on both tracks with a 5% processing fee. Tenures range from 3 to 18 months. See the pricing section above for full details.

Is this online or offline?

Primarily online via live, instructor-led sessions - real-time interaction from the comfort of your home. The campus immersion at IIT Roorkee is an in-person event.

What if I miss a session?

Every class is recorded and available for replay. We also generate an AI-powered summary of each session so you can catch up quickly.

How is this different from Coursera or Udemy?

This is a live, synchronous, instructor-led program - not a MOOC. You solve problems, invent solutions, and defend your thinking in real time. The goal is not to finish a course; it is to become a different kind of thinker.

I already know Python. Is this still valuable?

Absolutely. This program teaches the art of algorithm design: how to break down any problem, reason from first principles, and build your own solution from scratch. Most experienced developers have never had to think this way.

What is the minimum age requirement?

Anyone who has completed 10th grade is welcome. There is no upper age limit.

What laptop do I need?

Any laptop that can run Zoom. No special GPU or high-end hardware required.

Will I get placement support?

Yes. We run a Placement Eligibility Test (PET) after the program. Students who clear it have their CVs shared with 400+ hiring partners. We also directly hire strong performers into our consulting arm.

How much time outside of class do I need?

We recommend 1–2 hours of practice daily outside weekend sessions. The more you engage with problems between classes, the faster you grow.

When does the next batch start?

The 30 May 2026 batch is sold out. The next batch starts 30 June 2026 (8–10 pm IST, Sat/Sun) and is enrolling now. A third batch begins 30 September 2026 (8–10 am IST, Sat/Sun).

Can I do this while in college?

Absolutely - that is exactly what this program is designed for. Classes run only on weekends, 2 hours each session. Your weekdays are completely free for college. Finish college with a portfolio and an IIT Roorkee certificate.

What if my exams clash with the schedule?

Every session is recorded. Around exam periods, reduce practice hours and catch up on recordings. Dedicated catch-up sessions are built into the schedule.

I'm in 1st year of college - which certificate will I get?

You will receive the IIT Roorkee Certificate (Certificate track). The PG Certificate is for those who have already completed a Bachelor's degree. Both tracks follow the exact same curriculum and campus immersion.

Will AI replace data scientists?

AI will replace people who use tools without understanding them. It will not replace people who understand how those tools are built. The explosion of AI has created a shortage of engineers who can build, fine-tune, evaluate, and deploy AI systems. This program is designed to make you one of them.

Why learn algorithms when AI can solve problems?

The problems worth solving are the ones AI can't yet solve on its own. Algorithm thinking is about developing the instinct to decompose any problem into solvable pieces. That instinct cannot be outsourced - and it is what separates engineers who get hired from those who don't.