Learning Objectives

Turn an ambiguous business situation into a structured, solvable problem the way a consultant does.

Topics

• Framing a business problem — the symptom vs. the real question
• The SMART framework applied to analytics scopes
• MECE — mutually exclusive, collectively exhaustive
• Root cause analysis — 5 Whys, issue trees
• Solution evaluation — effort, impact, data availability
• Proof-of-concept thinking
• Masterclass: structuring client problems live

Hands-on:

• ChatGPT (market, competitor, and stakeholder research)
• Git & GitHub

Project - The MECE Correction (Git-tracked)

You’re given AI-generated issue trees for a business problem. Test them against MECE and find where they fail — AI trees routinely overlap branches or miss categories entirely. Commit history: raw AI tree → your MECE audit notes → corrected tree, each with an AI-assisted commit message. 

Learning Objectives

Run a complete analysis in Excel — clean, transform, summarize, visualize — and build pivot driven dashboards that answer a stated business question.

Topics

• Slicing and dicing data
• Complex functions and nested logic
• Text functions and cell referencing
• VLOOKUP / HLOOKUP / XLOOKUP
• INDEX-MATCH
• IF, nested IF, and IFERROR
• SUMIFS, COUNTIFS, AVERAGEIFS
• Data validation and conditional formatting
• Named ranges
• Pivot tables and pivot charts
• Data linking across sheets and workbooks
• What-if analysis, Goal Seek, and scenario modelling
• Chart formatting
• Basic dashboard logic
• AI-assisted formula generation and data-cleaning logic

Hands-on:

• Excel
• ChatGPT / Claude

Project - Sales Performance Dashboard

Clean a raw sales dataset and build a pivot-based dashboard that answers a few key business questions.

Learning Objectives

Build production-quality reports end to end — modelled correctly, written in DAX, ready to share.

Topics

• Power BI overview — where it fits against Excel and Tableau
• Data preparation in Power Query
• Data modelling — relationships and cardinality
• DAX fundamentals for business metrics
• Report building and layout
• Power BI Copilot — AI-assisted DAX and visual suggestions inside the tool itself
• Dashboard best practices

Hands-on:

• Power BI Desktop
• Power BI Copilot
• ChatGPT / Claude (AI-assisted DAX generation, storytelling support)

Project - Store Underperformance Dashboard

Build a dashboard from raw retail sales data to find why certain stores are missing targets, with AI-generated DAX measures verified before use.

Learning Objectives

Translate your Power BI skills to Tableau and broaden employability at Tableau-first organizations.

Topics

• Getting started with Tableau
• Charts and plots
• Hierarchies, grouping, and treemaps
• Calculated fields and functions
• Dashboards and storytelling
• LOD (Level of Detail) expressions
• AI-assisted calculated-field generation and chart-type suggestions

Hands-on:

• Tableau
• ChatGPT / Claude (calculated-field generation, chart-type prompting)

Project - Fresh Dataset Dashboard

Build a standalone Tableau dashboard on a new business dataset, with AI-assisted calculated fields verified before use.

Learning Objectives

Write practical Python for the analysis Excel can’t handle, and pair-program with AI while keeping full ownership of the code.

Topics

• Data types and operations
• Data structures — lists, dicts, tuples, sets
• Control structures — loops, conditionals
• Functions — writing your own, and important built-ins (map, filter, lambda, etc.)
• Error handling — try/except/finally
• File handling — reading and writing files
• The yield keyword and generators
• Decorators
• What an API actually is, and how a request/response works
• Calling AI directly from code — Claude/OpenAI/Gemini APIs, instead of only chatting
• Intro to Flask — building a tiny API of your own to serve results
• AI-native coding in Cursor — agent-assisted edits inside the same file

Hands-on:

• Python
• Flask
• Cursor
• Claude API / OpenAI API
• Streamlit
• Supabase
• Vercel
• Git & GitHub

Project - Expense Tracker

Build an expense tracker with a Flask backend API, a Streamlit UI, Supabase as the database, deployed live on Vercel.

Learning Objectives

Explore a dataset properly before drawing conclusions from it, then back those conclusions with statistics you can defend under challenge.

Topics

• EDA in Python
• • NumPy for numerical work
  • Pandas — reading, joining, grouping, reshaping real datasets
  • Matplotlib for visualisation
  • Data sourcing and ERP basics
  • Data cleaning as an analytical decision, not a preprocessing chore
  • Missing values and outliers — spotting and handling them
  • Univariate and bivariate analysis
  • Derived metrics — building the variable the question actually needs
  • AI as an EDA thought-partner — “what should I even check in this dataset?”
• Statistics
• • Random variables and probability distributions
  • Central Limit Theorem
  • Confidence intervals
  • Normal and t-distributions
  • Hypothesis testing
  • Type I / Type II errors
  • A/B testing — design, run, interpret

Hands-on:

• Python (NumPy, Pandas, Matplotlib)
• ChatGPT / Claude (EDA thought-partner, statistical explanation)
• Git & GitHub

Project - Zomato Data Analysis

Full EDA and statistical analysis on the Zomato dataset — cleaning, univariate/bivariate exploration, and at least one defensible hypothesis or A/B-style read.

Learning Objectives

Query relational databases with confidence and understand warehouse fundamentals well enough to work alongside data engineers.

Topics

• Database design and data warehouse basics
• Data modelling for analytics
• SQL querying fundamentals — SELECT, WHERE, GROUP BY, ORDER BY
• Joins and set operations
• Nested queries
• Window functions
• Solving business problems through data modeling
• Text-to-SQL — plain English question in, AI writes the query, you verify it
• AI-assisted query generation and refinement

Hands-on:

• MySQL
• Vanna.ai / ThoughtSpot Sage (text-to-SQL)
• ChatGPT / Claude (query generation, refinement)
• Git & GitHub

Project - Business Query Set

A business database and a list of real questions to answer purely in SQL — joins, nested queries, aggregations, and window functions.

Learning Objectives

Build and interpret core models, choose the right technique for the business question, and move from predicting what happens to recommending what to do.

Topics

• What is machine learning
• Supervised vs. unsupervised learning
• Classification vs. regression
• Overfitting and underfitting
• Classification algorithms — logistic regression, decision trees
• Regression algorithms — linear regression
• Time-series forecasting
• Demand forecasting
• What is NLP?
• What is an embedding, and the types of embeddings
• Evaluating ML models

Hands-on:

• Python (scikit-learn)
• ChatGPT / Claude (model-code generation, explanation)
• Git & GitHub

Project - Support Ticket Triage

Build a classification model to predict ticket priority/routing, then use embeddings to catch tickets that were tagged inconsistently by humans.

Learning Objectives

Convert finished analysis into a board-ready recommendation that changes what someone decides.

Topics

• The five patterns of insight — what analysis can actually surface
• Analysis approach and insight demonstration
• The pyramid principle — answer first, support second
• Logical flow — building an argument a skeptical executive can’t derail
• Communicating quantitative and qualitative insight together
• Visual design principles for decision-making, not decoration
• Storyboarding a readout before building a single slide
• AI-assisted summarization and documentation

Hands-on:

• ChatGPT / Claude (summarisation, insight writing, documentation)
• PowerPoint / Google Slides

Project - Three Audiences, One Insight

Take one finished analysis and turn it into three different readouts — for a CFO, an ops lead, and an external client — same underlying insight, different framing for each.

Learning Objectives

Connect a few steps together so AI can complete a small job automatically from start to finish, with no manual work in between.

Topics

• What workflow automation is, and how it’s different from writing a script yourself
• What an AI agent is, in plain terms
• Introduction to n8n — nodes, triggers, connections
• Common triggers — new file, new form response, scheduled time, webhook
• Chaining steps — trigger → AI processing → action, and multi-step chains beyond that
• Conditional logic in a workflow — branching based on what the data says
• Connecting AI into a workflow — where the “AI step” sits and what it needs as input
• Error handling in automations — what happens when a step fails

Hands-on:

• n8n
• Google Drive
• Claude API / OpenAI API
• Git & GitHub

Project - Self-Updating Dashboard

A source file in Google Drive keeps getting updated with new data. Build an n8n workflow that detects the change, cleans and processes the new data with AI, and automatically refreshes a dashboard — no manual re-uploading, no rebuilding it by hand.

AI used across the full workflow — research, code, queries, narrative. Every AI touchpoint, verification performed, and error caught is documented in a mandatory appendix, submitted alongside both capstones.

Learning Objectives

Execute a full analyst-consultant engagement twice, and present client-style readouts to a review panel with mentor involvement.

Topics

• Capstone A: Indian-market dataset (retail/fintech/quick-commerce — pricing, churn, or demand)
• Capstone B: International dataset (global e-commerce or supply chain)
• End-to-end: problem framing → data → analysis → model → dashboard → recommendation

Hands-on:

• Everything from Modules 1–11
• Git & GitHub

Project - AI-Usage Appendix

AI used across the full workflow — research, code, queries, narrative. Every AI touchpoint, verification performed, and error caught is documented in a mandatory appendix, submitted alongside both capstones.