AI Fluency
Walk in with a degree. Walk out with proof.
Six weeks, one project of your own. You build it with AI, step by step, and defend it in front of the room after every five classes. Leads to the Certificate in AI Fluency Foundations.
- Duration
- 45 days: 36 sessions over six weeks, then 3 days of individual feedback and certificates
- Sessions
- 30 classes of 60 minutes + 15 minutes hands-on, and a demo session after every five classes
- Schedule
- Monday to Saturday, Sunday off
- Batch
- Maximum 10 students
- Where
- Kidwai Nagar, Kanpur (offline, in person)
- Fee
- ₹4,999 founding batch price (MRP ₹7,999)
- Prerequisites
- None. No coding, no maths. Laptop required.
- Language
- English on screen, Hinglish in the room
Built for graduates who need proof
Designed for
Graduation students and recent graduates looking for a job, from any stream: B.Tech (any branch, CSE included), BCA, MCA, B.Com, BBA, BA or B.Sc.
Your degree looks the same as thousands of others, and employers now expect people to use AI at work from day one. In six weeks you build proof: a live project of your own, defended in front of a room, with a portfolio and an ATS-ready resume that point to it.
Also welcome: early-career professionals who want AI in their current job, and young people working in a family business.
Not for
- People who want to write ML code or become ML engineers
- Anyone expecting a job guarantee
- Anyone who wants to watch a list of tools without building anything
What you can do at the end
Every capability is shown on your own project, not just seen in class.
Explain how AI works
Explain how an LLM works, why it hallucinates and when not to trust it, in two minutes, to someone who does not know AI.
Get the output you want
Write a CO-STAR+ prompt for an unseen task that gets a usable result without the model guessing.
Check AI’s work
Find the errors in an AI output and decide how far a task can be trusted to AI.
Build and change a website
Turn a written requirement into a live website without code, then change it from a written spec.
Build an assistant and an automation
Build an AI assistant on your own documents, and one automation that runs without you.
Learn any new AI tool in 20 minutes
Report what an unfamiliar tool does, what it costs and how it fails, today’s tools or next year’s.
Present and defend your work
Present in English under questions from the room, including a live change request.
One project. Reviewed. Defended.
One project, your own problem
In the first week you choose a domain you care about and one problem in it. Every class after that builds on it: a requirement document, a website, a campaign, an assistant and an automation.
A demo after every five classes
Six demo sessions. You present what the block built, explain your decisions, answer questions from the room and say where AI should not be used in your project.
Reviewed, not graded
Every homework and every demo presentation is rated 1–5 and returned with written feedback, so you always know what to fix next. Both instructors review your work.
Certificate
Certificate in AI Fluency Foundations for at least 80% attendance. Below that, a Letter of Course Completion. No test, no grades.
The eight modules
Understand AI → specify work → build → create → analyse and automate → protect and get hired. After each module, you can say:
AI Foundations
“I know what AI is, what it is not, and I can work inside a modern assistant.”
AI & Information
“I know every kind of data I will meet, what a machine sees in it, what is sensitive, and where it goes.”
Inside the Model
“I can explain how an LLM works and how it was trained, and I know when not to trust it.”
AI Interaction
“I can specify work so the model never guesses, and I can make AI do the asking.”
AI for Building
“I can go from a requirement to a pitch, a brief, a live website and an automation.”
AI for Creation
“I can make images, video, voice and a full communication plan on purpose, for any audience.”
AI for Analysis, Assistance & Automation
“I can read business data, build an assistant that knows my documents, and wire it into a workflow.”
AI for Work & Career
“I can defend my work, get past an ATS and an AI interviewer, use AI on day one of a job, and know where AI should not be used.”
Session by session
Five classes, then a demo session, in six blocks. Each class is 60 minutes plus 15 minutes of hands-on work on your own project. Click a block to expand.
- Session 01
Welcome, and What AI Actually Is
An AI agent completing a real task from one sentence · AI before and after November 2022 · the course and your project · what AI is, in one line · three real tasks: explain, make, decide · the follow-up habit
You build: Three tasks done in ChatGPT, each with a follow-up; your one-line definition of AI
- Session 02
Work Inside a Modern AI Assistant
Memory and custom instructions · working with PDFs, images and spreadsheets · voice mode and search · prompting in Hindi and Hinglish · voice mode as a spoken-English coach · Projects, canvas and Study Mode · the 20-minute method for learning any new AI tool
You build: Your personal ChatGPT setup, a course Project, the tool-learning checklist
- Session 03
Data 1: Data You Can Read
What data is and how it drives decisions · structured, semi-structured and unstructured data · hands-on with txt, pdf, docx, csv, xlsx and json files · turning messy text into a clean table
You build: A classified file set and an extraction table
- Session 04
Data 2: Data a Machine Has to Decode, and What Is Personal
Images, video and audio as numbers · scanned vs digital PDFs · personal data and how to spot it · who owns data and AI-generated media · NotebookLM across your own files
You build: Your NotebookLM course notebook
- Session 05
Where Your Data Goes
The journey of a prompt from your laptop to the model and back · where data is stored · what happens to what you paste · why identifiers such as mobile and Aadhaar numbers make data personal
You build: A journey diagram and your personal “never paste” list
- Demo 1
Problem Selection
Present and defend your domain and problem · a short baseline task done unaided, repeated at the final showcase · your AI field report: five real tasks, what worked, what failed · AI in your language: Hinglish prompting, Bhashini, Sarvam · India’s AI labs
You build: Your problem statement
- Session 06
How an LLM Works
Tokens and how text becomes numbers · next-word prediction · the context window and losing context · hallucination · cost, speed, free vs paid · open vs closed models
You build: Three hallucinations found and verified
- Session 07
How an LLM Is Trained
Data collection and pre-training · fine-tuning and human feedback, and who labels the data · safety, refusals and evaluation · bias and the urge to please · reasoning models · what AI should not do: three questions and the trust ladder
You build: One bias, one refusal and one reasoning comparison; your own tasks placed on the trust ladder
- Session 08
Prompting 1: CO-STAR+
Why prompting is a job · a question vs a prompt · the CO-STAR+ framework · hard constraints · how information changes the output · why longer is not better
You build: The first project prompt in your prompt library
- Session 09
Prompting 2: Show, Don’t Tell, and Set the Rules
Examples in the prompt for structured output · why one example breaks and a few beat a smarter model · label bias · system instructions · hands-on in Google AI Studio
You build: An extraction prompt and an assistant system instruction
- Session 10
Prompting 3: Let AI Do the Asking
AI that writes prompts for you · interviewer mode: AI questions you until it can write the document · Deep Research · prompt chaining · when to ask for step-by-step reasoning
You build: Your requirement document, started
- Demo 2
Requirement Build Session
Requirement document finished with the instructor · prompt clinic · demo at the end of the session
You build: Your requirement document
- Session 11
Presentations That Work
Common industry deck types, with a sample of each · CO-STAR+ prompts for decks · Gamma, Napkin and Slides with Gemini compared · fixing one slide by hand
You build: Your pitch deck
- Session 12
Plan Your Website
What a website is made of: HTML, CSS and JavaScript · what a URL fetches · static vs dynamic sites, builders vs code · what an API is · a site brief written by AI interviewing you · a first draft page in Google AI Studio Build
You build: Site brief, first draft page, GitHub account
- Session 13
Build, Compare and Publish
What one prompt can build · building together, step by step · the index.html file explained · publishing from a builder and from code with GitHub Pages · SEO basics
You build: Your live project website
- Session 14
Make AI Run a Task Without You
Claude as today’s example · chat vs Cowork · Projects · schedules and connectors · Claude in Chrome
You build: One scheduled automation
- Session 15
Change a Live Site from a Written Spec
Why a written spec beats chatting · Google AI Studio Build from a spec · AI that edits files, demonstrated with Gemini CLI · recovering when it goes wrong
You build: Site v2 live, spec updated
- Demo 3
Website Demos, Portfolio Build, the AI Tool Landscape
Present and defend your website · build your personal portfolio site · the AI tool landscape: which tool for what, and switching when one hits a limit
You build: Your portfolio site
- Session 16
How Image Models See and Learn
How image models learn · why images are harder than text · text models vs vision models · trending image commands as practice in a precise visual spec · the image prompt collection · from a paper sketch to a one-page website
You build: Three prompts from the collection adapted to your project
- Session 17
Image Prompting and Marketing Ads
Image controls beyond CO-STAR+ · what makes an ad image work · choosing a tool: ChatGPT Images, Nano Banana, Ideogram, Canva · three ads built step by step · explainer images when your project is not a business
You build: Three ad images and a hero image for your site
- Session 18
Video Generation and Ad Videos
Google Flow: prompts, frames, extending, camera · camera language · the structure of a 20–30 second ad · launch or explainer videos for projects that are not businesses
You build: Your project ad video
- Session 19
Audio: Voice and Dubbing
Text to speech and voice cloning · dubbing, translation and sound effects · consent · deepfakes and voice-clone scams: how they are made and how to spot them
You build: A voiceover and a dubbed clip
- Session 20
The Marketing Manager’s Mind
What a campaign is, and why every project has an audience · how a marketing manager thinks · four context files: product, audience, voice, offers · chaining prompts from positioning to captions, scripts and emails · good vs bad captions and hooks · assembling in Canva and Sheets
You build: Your context files and a two-week content pack
- Demo 4
Campaign Demos
Present your ads and content pack · Hindi and regional-language versions of a campaign · the AI gig economy: packaging what you can now make, selling a website to a local business, pricing, finding a first client
You build: Your campaign, presented
- Session 21
Data Analytics 1: What Happened?
What analytics is, before and after AI · the analyst’s loop · the funnel from spend to orders · CTR, CPC, conversion, cost per lead, CAC, ROAS · questions answered on a shared dataset with ChatGPT and Gemini in Sheets, one number checked by hand
You build: A funnel sheet and your first three answers
- Session 22
Data Analytics 2: Why, and What Next?
Diagnostic and predictive questions · choosing the right chart · the DECIDE frame and a dashboard · the drop, the budget, products sold together, next month
You build: A one-page data report
- Session 23
Build an Assistant from Your Own Material
From Eliza to menu bots, intent bots and LLM bots · custom GPTs and Gems · building from your system instruction and context files
You build: Assistant v1, tested on ten questions
- Session 24
RAG: The Open-Book Exam
Why you cannot paste everything · chunk, embed, retrieve, answer · no-code RAG with NotebookLM, Gem files and custom GPT files · fixing wrong answers
You build: Assistant v2
- Session 25
Ship a Chatbot, Wire an Automation, Meet Agents
A no-code chatbot on your site · an automation from form to sheet to WhatsApp or email · agents: answer, automate, decide
You build: A public chatbot and a live automation
- Demo 5
Intelligence Demos and WhatsApp Deployment
The ladder so far: generate, work on information, assist, perform a workflow · present your report, bot and automation · your bot on WhatsApp Business · on request: a mobile app with AI
You build: Your bot on WhatsApp
- Session 26
Adversarial Prompting: Attack and Defend Your Bot
Prompt injection vs jailbreak · why attacks work · hidden instructions in resumes and role-play tricks · defences applied and re-tested
You build: A hardened bot
- Session 27
Getting Hired 1: ATS, Resumes, the Job Description
How an ATS reads a resume · keywords vs stuffing, and formats · the job description as the spec, and gap analysis · a tailored resume and an ATS check · your project and portfolio as evidence
You build: An ATS-ready resume
- Session 28
Getting Hired 2: Interviews, LinkedIn, Portfolio Story
Spoken mock interviews in ChatGPT voice mode · AI interviewers: how companies use them and how to face one · LinkedIn from your resume · your portfolio story: the requirement, what AI did, what failed, the fix
You build: A two-minute pitch and an updated LinkedIn
- Session 29
Your First Week at Work with AI
Email drafts and thread summaries · meeting notes · formula help in Excel and Sheets · Microsoft Copilot in Word, Excel and Teams, demonstrated · translation and rewriting · what company AI policies usually say
You build: A day of work tasks done with AI
- Session 30
Privacy, the Future, and Rehearsal
Where AI should not decide: hiring, health, money, exams, other people’s data · what to paste and what never to · India’s DPDP Act, briefly · AGI, ASI and the future of work, without hype · full showcase rehearsal
You build: A data policy on your site; your showcase ready
- Demo 6
Final Showcase
Present the whole project, with one live change request · the baseline task repeated unaided, shown beside your Demo 1 result · Certificate in AI Fluency Foundations
You build: Your certificate
Running through the course
Seven threads that no single module owns.
Your project
Chosen in the first week, grown in every class, shown in every demo session and presented at the final showcase.
Communication and confidence
You present and defend your work in every demo session, and practise spoken English with AI voice mode from Session 02.
What AI should not do
Three questions (what if it is wrong, can I verify it, would I be comfortable if people knew) and the trust ladder, applied to your own project in every demo.
The 20-minute method
How to learn any new AI tool: what it does, what it takes, what it costs, how it fails, how to verify. Taught in Session 02, used in every demo session.
Your language and Indian AI
Prompting and voice in Hindi and Hinglish, and AI built in India, from Session 02 onward.
Feature of the day
A short look at one product feature, included only when one fits the session’s topic.
From the industry
Each module carries one real, anonymised story from the instructors’ work: how it is actually done, what went wrong and what fixed it.
What we deliberately do not teach
| Not taught | Why |
|---|---|
| Writing code by hand | You build working results by generating code and copying it, not writing it. Course 2 is for those who want to code. |
| Mathematics and ML theory | A full mental model is enough to use AI well and judge its output. |
| Training or fine-tuning models | Not part of a user’s first job. |
| Paid tools for your work | You must be able to repeat everything after the course at no cost. |
| Lists of “100 AI tools” | Lists go stale in months. The 20-minute method lets you learn any tool yourself. |
| Prompt hacks and magic phrases | Tricks tied to one model break with the next release. One framework works everywhere. |
| More than one prompting framework | One method practised for six weeks beats five methods seen once. |
| A job guarantee | The course builds evidence: a project, a portfolio, a resume and interview practice. Employers decide. |
Today’s tools
These are today’s tools (October 2026). The course teaches capabilities and the 20-minute method for learning any new tool, so you are ready when the tools change. All student work runs on free tiers; paid tools appear only in instructor demonstrations.
| Capability | Today’s tools |
|---|---|
| Work inside a modern AI assistant | ChatGPT; Gemini and Claude as alternatives |
| Study across your own files | NotebookLM, ChatGPT Study Mode |
| Specify work and set system instructions | ChatGPT, Google AI Studio |
| Make a presentation | Gamma, Napkin, Slides with Gemini |
| Build and publish a website | Google AI Studio, GitHub Pages |
| Make AI run a task without you | Claude Projects, schedules and connectors, Claude in Chrome |
| Change a live site from a written spec | Google AI Studio Build; Gemini CLI (demonstrated) |
| Make images on purpose | ChatGPT Images, Nano Banana, Ideogram, Canva |
| Make video | Google Flow |
| Read business data | ChatGPT, Gemini in Sheets |
| Build an assistant from your own material | Custom GPT, Gem, NotebookLM |
| Ship a chatbot and an automation | Chatbase, Botpress or Tidio; WhatsApp Business |
| Use AI on day one of a job | Gemini in Meet, Otter, Microsoft Copilot |
Your instructors
The two decide per session who teaches it. Both review every homework and every demo presentation.
Arpit Tiwari
Founder · Senior Data Scientist
Senior Data Scientist with 5.5+ years of Fortune 100 experience. Leads the course.
Kajal
Co-founder · Senior AI Engineer
Senior AI Engineer who also teaches and mentors in her work. Co-teaches the course.
Ready to become AI‑fluent?
Fixed-date cohorts · maximum 10 students per batch · founding price locks at enrolment.
36 sessions over six weeks · a demo after every five classes
- One project of your own, from idea to live website
- Every homework and demo reviewed, with written feedback
- Certificate in AI Fluency Foundations
Pro-rata refund on the offline cohort · see Terms · Download Full PDF