Six sessions, ~10 hours each. Each session ends with a working artifact you ship to your public profile — a prompt library, a real management deck, a live dashboard, a 1-page insights memo, an automation, and a full capstone solution.
Three AI assistants are now better than any colleague at writing, translating, and summarising — but only if you talk to them properly. This session you stop typing into ChatGPT like a Google search and start treating it like a junior teammate. By the end of the week, the routine writing tasks that ate two hours a day take twenty minutes.
Every example in this session is a real Indian-workplace task — replying to a vendor on WhatsApp, drafting a polite escalation email to a defaulting client, summarising a 90-minute meeting that ran in Hinglish, translating a circular into Hindi for the shop floor. The goal is to leave Session 1 with twenty prompts that map to your actual job, not generic templates.
Learning objectives
- ›Pick the right AI tool — ChatGPT, Claude, or Gemini — for each kind of task
- ›Write prompts that follow the role + context + constraint + format pattern
- ›Build a personal prompt library of 20+ tested prompts for your routine work
- ›Turn meeting recordings into action-items lists with owners and dates
Lessons
- ·Three assistants — and when to use which60m
- ·Prompts that don't waste your time75m
- ·Build your prompt library + meeting workflow90m
Portfolio piece — My AI Prompt Library + a real meeting workflow
The single artifact that proves you've moved from "I've heard of ChatGPT" to "I use AI every day at work." Two parts:
1) A Google Doc with 20+ tested prompts — each with a title, a when-to-use, and a fillable template — organised under Emails, Documents, Translation, Analysis. Every prompt has been run at least once on a real example from your job.
2) One real meeting summary you generated this week using the recording → AI → action-items workflow. Share the action-items table only (not the full transcript). Anonymise names if needed.
Set the Google Doc share setting to "Anyone with the link can view" before submitting.
A management presentation that used to take a day now takes 25 minutes. Same for SOPs, proposals, vendor briefs, FAQs. This session you stop opening blank slides and blank docs — and start describing what you want in one paragraph. Output — a real management deck and a real SOP, end-to-end with AI, that you can show your manager tomorrow.
Every example is a document your office actually produces — a quarterly review for the management committee of an SME, a vendor onboarding SOP, a proposal for an Indian client, an HR policy in plain Hindi. We use Gamma free, Google Slides + AI, Microsoft Copilot if you have it, Tome, and ChatGPT/Claude as the ghostwriter behind all of them.
Learning objectives
- ›Generate a 10-slide management deck from a one-paragraph brief in under 25 minutes
- ›Write SOPs, proposals, and FAQs by describing the process in plain English
- ›Use Gamma, Tome, Copilot for PowerPoint, and Google Slides + AI on free tiers
- ›Edit AI-generated drafts efficiently — keep the structure, fix only what matters
Lessons
- ·From one paragraph to a 10-slide deck75m
- ·SOPs, proposals, and FAQs at speed75m
- ·Ship a real deck and a real document75m
Portfolio piece — A real deck and a real document
Two artifacts you'd actually use at work:
1) A 10-slide deck made in Gamma (free) from a one-paragraph brief, exported as PDF or shared as a Gamma link. Pick the scenario that fits your role — sales pitch, monthly review, onboarding deck, quarterly numbers, campaign report, or company analysis.
2) One real workplace document — SOP, proposal, FAQ, or HR policy in Hindi — generated using the matching prompt template and finished with the 3-pass edit. Google Doc or PDF.
Both must be the real thing, not a practice exercise. If you'd be embarrassed to show it to your manager, redo it.
Indian offices run on Excel. This session you stop writing formulas from memory and start describing what you want in plain English to ChatGPT — VLOOKUP, XLOOKUP, SUMIFS, pivot tables, the works. Then you take that cleaned spreadsheet and turn it into a live Looker Studio dashboard your manager can actually open on their phone. Free tools all the way through.
We work with the kinds of spreadsheets every Indian SME has — sales registers with party-wise totals, inventory with reorder levels, GST receivables aging, attendance, expense claims. We connect them to Looker Studio (free, Google) for the dashboard. By the end of the session you'll have built a real dashboard for a real Indian SME workflow — from raw data to phone-friendly visual, no code, in under 3 hours.
Learning objectives
- ›Use ChatGPT to write Excel and Google Sheets formulas by describing the task in plain English
- ›Build pivot tables and dynamic charts without remembering syntax
- ›Connect a Google Sheet to Looker Studio and ship a live, shareable dashboard
- ›Recognise the most common Indian-SME data patterns (sales register, GST aging, inventory, attendance)
Lessons
- ·Stop memorising formulas — describe them75m
- ·Pivot tables, charts, and clean data75m
- ·Ship a live dashboard in Looker Studio90m
Portfolio piece — A live Indian-SME dashboard
A real, phone-friendly Looker Studio dashboard connected to a Google Sheet, built for a real Indian SME use case — sales, inventory, attendance, cash flow, or marketing leads.
Pick the use case that matches your job (or imagined first job). Source data can be real (anonymise sensitive columns), a public Indian-business Kaggle dataset, or a sample you generate that mimics the real shape.
Submit:
1. Looker Studio share URL (anyone with link can view)
2. Source Google Sheet share URL
3. A screenshot of the dashboard on phone view (1080 × 1920)
Last session you built a dashboard that answers the questions you anticipated. This session you handle the questions you didn't — by uploading a CSV to ChatGPT or Claude and asking in plain English. "Which 5 customers slipped the most this quarter compared to last?" gets answered in 30 seconds, with a chart. No SQL, no Python, no pivot tables. The catch — you have to know how to verify the answer, because AI confidently makes up plausible numbers.
We work with the kinds of ad-hoc questions managers actually ask in Indian SMEs — "why did sales drop in Pune last month?", "are we losing more on COD vs prepaid orders?", "which products have the highest return rate from south India?" The answers come from CSVs you upload, no IT department needed.
Learning objectives
- ›Use ChatGPT Advanced Data Analysis (free tier limited) and Claude file-upload to answer business questions in natural language
- ›Verify AI-generated analysis by spot-checking the underlying numbers, not just trusting the chart
- ›Recognise the three modes where AI data analysis goes wrong (hallucinated numbers, wrong groupby, leaking averages)
- ›Combine dashboard + ad-hoc analysis into a reusable analyst workflow
Lessons
- ·Upload a CSV, ask in English, get answers75m
- ·When AI gets the analysis wrong60m
- ·Ship a real analysis with insights75m
Portfolio piece — A 1-page insight memo from real data
A 1-page insight memo answering one real business question, using a real (or realistic) CSV as the source. Format: question → finding → 3 verified numbers → likely cause → recommendation → caveats → method footnote.
This is the format Indian SME managers actually want from analysts. Decks lose attention. Memos earn follow-up questions.
Submit:
1. The 1-page memo (Google Doc, "anyone with link can view")
2. The source CSV (or link to where you got it — Kaggle URL is fine)
3. Optional: a screenshot of one supporting visual
This is the session everyone in your office will eventually ask you to teach them. Zapier and n8n connect the apps your office already uses — Gmail, Sheets, WhatsApp, forms, Razorpay — and AI sits inside the automation, classifying, summarising, extracting. By the end you'll have an invoice OCR pipeline (photo → extracted GSTIN/amount → logged to Sheet) and a custom GPT chatbot trained on your company's documents. Both real, both shippable.
Every automation we build solves a real Indian-SME pain — the accounts team manually entering 200 GST invoices a month, the sales team copy-pasting WhatsApp orders into Excel, HR answering the same 20 leave-policy questions every week. The free tier of n8n self-host is mandatory; Zapier free is for the simpler ones; ChatGPT custom GPTs handle the chatbots. No coding.
Learning objectives
- ›Build trigger-action automations in Zapier (free tier) and n8n (self-hosted free)
- ›Insert ChatGPT/Claude into automations to classify, summarise, and extract structured data
- ›Build an invoice OCR pipeline that reads photos and extracts GSTIN, amount, vendor name
- ›Build a custom GPT chatbot trained on company documents (HR policy, product catalog, SOPs)
Lessons
- ·Trigger → action — building your first automation75m
- ·AI inside the automation — invoice OCR pipeline90m
- ·Custom chatbots without code (and your portfolio piece)90m
Portfolio piece — One automation + one chatbot
Two real artifacts your office (or a real Indian SME you're targeting for your first job) would actually adopt.
Artifact 1 — End-to-end automation with AI inside (built in Zapier, Make, or n8n self-hosted): pick from invoice OCR, resume screener, email triage, meeting summary auto-emailer, customer feedback classifier, or WhatsApp order parser. Trigger → AI step → action.
Artifact 2 — A custom chatbot trained on real (or realistic) documents (Claude Projects, custom GPT, or Chatbase): HR Policy Bot, Sales Assistant Bot, SOP Bot, or Customer Support Bot.
Submit both as a single portfolio entry.
Session 6
Capstone — pick your vertical, ship one solution
3 lessonsAssessment · ≥70%CapstonePortfolio piece
One end-to-end AI-powered solution for one Indian-SME workflow in the vertical you'll work in. Real data, real automation, real chatbot, real dashboard, real business case ("saves 8 hrs/week = ₹2.4L/year"), real Loom demo. This is the artifact you put on your CV. Recruiters and SME owners can click your portfolio link and see something working — not a quiz score.
Pick one of five verticals — Accounts/Finance, Sales/Marketing, HR/Admin, Operations/Supply chain, Healthcare/Education admin. The capstone is built for an actual Indian-SME pain in that vertical. Five worked examples below; pick one or invent your own.
Learning objectives
- ›Combine all six skills (prompts, slides, Excel/dashboard, analysis, automation, chatbot) into one solution
- ›Write a business case that translates time-savings into rupees
- ›Position yourself as 'AI-powered Operations Analyst', not 'ML engineer', in your CV and interviews
- ›Ship a portfolio piece that earns interviews on its own
Lessons
- ·Pick your vertical and define the problem60m
- ·Build the capstone240m
- ·Business case + CV positioning + ship90m
Portfolio piece — Workplace Pro Capstone
ONE end-to-end AI-powered solution for ONE Indian-SME workflow in ONE chosen vertical (Accounts/Finance, Sales/Marketing, HR/Admin, Operations/Supply chain, or Healthcare/Education admin).
Combine at least 3 of the prior 5 sessions' skills: prompts, slides/docs, dashboards, ad-hoc analysis, automation, chatbots.
Submit:
1. A 4-7 minute Loom video walking through problem → solution → live demo → before/after numbers
2. A 1-page business case (Google Doc, "anyone with link") with the time/rupee math
3. The Looker Studio dashboard link (anonymised)
4. The chatbot share link OR a recording of it answering 5 real questions
5. Screenshots of the n8n/Zapier workflow
6. A 100-word "what I learned" reflection in the description
This is the artifact that goes on your CV. Treat it like one.