Lab
Benchmark suite
Time and memory for list vs dict vs heap vs graph on the same task.
Data Structures & Algorithms
The structures and algorithms AI systems actually sit on: arrays, hashes, graphs, heaps, and complexity you can measure — taught in Python, aimed at engineering work.
Created by Baljeet Dogra
Course Objectives
- Read a complexity bound and know what it costs on a 10-million-row corpus.
- Pick a structure because of access pattern, not because a tutorial used it.
- Implement graph search, heaps and hashing in Python you can test.
- See where RAG, schedulers and batching are just these structures in production clothes.
What you will produce
Build
Mini search index
Inverted index plus a heap for top-k — the skeleton of retrieval.
Write
Complexity memo
One page: why this structure, what you rejected, where it breaks.
6 weeks curriculum
Expand a part for the syllabus. Content stays searchable when closed.
01 Complexity you can feel Week 1
Big-O in practice. Best, average, worst. Why an O(n²) chunker dies on a 50k-page corpus.
- Asymptotics
- Amortised cost
- Profiling in Python
- When constant factors win
02 Sequences and hashing Week 2
Arrays, lists, stacks, queues, dicts and sets. Collisions, load factor, and why your cache is a hash map.
- Lists vs arrays
- Hash tables
- Sets
- LRU as a structure
03 Trees and heaps Week 3
Binary trees, heaps, priority queues. Scheduling jobs and top-k retrieval are the same shape.
- Heap operations
- Priority queues
- Tree walks
- Top-k
04 Graphs and search Weeks 4–5
BFS, DFS, Dijkstra. Agent graphs and knowledge graphs are not a different subject.
- Representations
- BFS / DFS
- Weighted paths
- Cycles
05 Sorting, selection, capstone Week 6
Sorts you should know, selection, and a capstone: an inverted index with measured complexity.
- Quick / merge / heap sort
- Selection
- Inverted index
- Write-up
Who this is for
Engineers before the AI stack
You write Python. You have not yet had to argue for a structure under load.
Career changers
You need DSA that shows up in interviews and in RAG, not contest puzzles for their own sake.
Prerequisites
- Python functions, lists and dicts
- Comfortable in a terminal
- Git at a basic level
Not a fit if
- Competitive-programming olympiad training
- A first course in programming — start with Python for GenAI
Related: Python for GenAI · Mathematics for AI
Questions
Is this LeetCode?
No. You will write real implementations and measure them. Interview-shaped questions appear, but the point is structures you reuse in AI systems.
Why six weeks?
Enough to implement, profile and write. Not a year of textbook chapters you will not open again.
Ready to start?
Six weeks from complexity to a measured mini-index. Create an account to enrol.
Enrol now