Part 2 · Lectures 11–20

Big Data Management
& Analytics

Ten sessions on non-relational data. We model, query, index and visualise one real ride-booking dataset in MongoDB, then rebuild the same dashboard in Redis — and argue about which was the right choice.

10
Lectures
5
Modules
15,000
Rows of data
90
Min / session
Part 1 Relational Foundations & SQL Lectures 1–10 · completed
L01–L10 Covered
Relational Modelling, SQL & the Data Warehouse
Normalisation, joins, transactions and analytical SQL — the baseline everything in Part 2 is measured against.
First half of semester
Module 1 Non-Relational Databases & MongoDB 101 Lectures 11–12
Lecture 11 Available
Introduction to Non-Relational Databases
Why relational hit a wall, vertical vs horizontal scaling, CAP, ACID vs BASE, and the four NoSQL families.
90 min · CLO1 Notes Slides
Lecture 12 Available
MongoDB 101 — The Document Model
BSON, collections, embedding vs referencing, schema design patterns, and loading your first Atlas collection.
90 min · CLO1 Notes Slides
Module 2 MongoDB 201 — Querying, Indexing & Aggregation Lectures 13–14
Lecture 13 Available
Querying Data with MongoDB
Full CRUD in mongosh, query and update operators, projections, sorting, and handling absent fields.
90 min · CLO3 Notes Slides
Lecture 14 Available
Indexing & the Aggregation Framework
B-tree indexes, the ESR rule, reading explain(), and building multi-stage pipelines that answer business questions.
90 min · CLO3 Notes Slides
Module 3 Data Visualisation with Atlas Charts Lectures 15–16
Lecture 15 Available
Atlas Charts & the Ride-Analytics Problem
A real business brief: 37.71% of bookings fail. Build the first dashboard and find out where the money leaks.
90 min · CLO4 Notes Slides
Lecture 16 Available
Dashboard Storytelling & the Recommendation
Cancellation deep-dive, the 17:00 insight, chart-choice discipline, and turning a dashboard into a decision.
90 min · CLO4 Notes Slides
Module 4 Key-Value Databases & Real-Time Analytics with Redis Lectures 17–18
Lecture 17 Available
Redis & the Key-Value Model
Strings, hashes, lists, sets, sorted sets, bitmaps and HyperLogLog — and why 1M users fit in 125 KB.
90 min · CLO2 Notes Slides
Lecture 18 Available
Building a Real-Time Analytics Dashboard
Compute-on-write instead of compute-on-read: a Node.js writer, a Redis key schema, and O(1) dashboard reads.
90 min · CLO2 Notes Slides
Module 5 Project Presentations & Course Wrap-Up Lectures 19–20
Lecture 19 Available
Project Presentations I
The full project brief, the technical checklist, the grading rubric, and the run sheet for presentation day.
90 min · CLO5 Notes Slides
Lecture 20 Available
Project Presentations II & Wrap-Up
Remaining teams, then the synthesis: compute-on-read vs compute-on-write, polyglot persistence, and where to go next.
90 min · CLO5 Notes Slides
Resources Course Materials Used throughout Modules 1–5
Dashboard Live
Uber Ride Analytics Dashboard
The finished reference dashboard — KPIs, status mix, hourly demand, vehicle revenue and cancellation reasons.
Built in Lectures 15–16 Open
Build guide Available
Rebuild It in Atlas Charts
The ten aggregation pipelines behind the reference dashboard, each with its Atlas chart type and encoding channels — build the same dashboard on your own cluster.
Lecture 15 lab Open
Reference Available
MongoDB & Redis Command Reference
Every mongosh and redis-cli command used in the course, on one searchable page.
Quick reference Open
Assessment Available
Assignments & Project Submission
Grading split, three assignments, the team project brief, rubrics and how to submit.
40% + 35% + 25% Open
Dataset Download
bdma.rides — 15,000 Ride Bookings
Delhi NCR, calendar year 2024. JSON for mongoimport, CSV for Charts, and a precomputed summary.
Seeded & reproducible JSON CSV