Starter template for the Python software project course. It gives your team every document you must submit — proposal, SRS, flowcharts, report, presentation outline, viva prep — as fill-in-the-blank templates, plus stubs for the AI module your project includes. Each section links to the learn.kevalabs.com guide that teaches the concept you need to fill it in.
You don't start from a blank page. You start from here.
- One team member clicks Use this template → Create a new repository (name it after your project, e.g.
cafe-order-engine). Make it private if you like, and add your teammates as collaborators. - Everyone clones it. New to Git? Read How Git works first.
- First assignment (do it now): find your project's pack in
docs/packs/, copy it todocs/project-pack.md, delete the other three packs, commit, and push. Congratulations — your repo is now your project's repo, and you've made your first team commit. - Work through the milestones below. Every document in
docs/tells you what to write and what to read.
| Path | What it is | When you need it |
|---|---|---|
docs/project-pack.md |
Your project's scope, data model, and AI formulation | Day 1, then constantly |
docs/proposal.md |
Proposal template | Milestone 1 |
docs/srs.md |
Software Requirements Specification template | Milestone 2 |
docs/flowcharts.md |
Flowchart + ER diagram starters (Mermaid — renders on GitHub) | Milestone 2 |
docs/report.md |
Final project report template | Milestone 4 |
docs/presentation.md |
Mid-term and final presentation outlines | Milestones 3 & 4 |
docs/viva-prep.md |
Questions the examiners will ask, with where to find answers | Before the viva |
labs/ |
Session worksheets — the follow-along labs, one per class | Every session |
ai/search.py |
Frontier-loop search skeleton (BFS done; DFS, greedy, A* are yours) | Cafe, Restaurant, Party Palace teams |
ai/rules.py |
Backward-chaining engine (engine done; the rules are yours) | Gym team |
src/ |
Your Django project lives here — you create it in the lab sessions | From the CRUD milestone on |
Internal — 60 marks
- M1 · Proposal (10 marks) —
docs/proposal.mdcomplete, topic approved by supervisor. - M2 · Requirements & design —
docs/srs.mdanddocs/flowcharts.mdcomplete. (Not separately marked, but the mid-term presentation is built from these.) - M3 · Mid-term presentation (20 marks) — working CRUD demo + AI module plan. Outline in
docs/presentation.md. - M4 · Pre-final submission & final presentation (30 marks) — working app with AI module integrated, report drafted, demo rehearsed.
External — 40 marks
- M5 · Project documentation (20 marks) —
docs/report.mdfinished and printed. - M6 · Final presentation (10 marks) — same demo, tighter.
- M7 · Viva (10 marks) — every member can answer everything in
docs/viva-prep.md.
These keep the project finishable in 45 hours. Breaking them needs supervisor approval, and the answer is probably no:
- Django + PostgreSQL, server-rendered templates only. No JavaScript frameworks, no REST APIs, no real-time features. (PostgreSQL setup is a one-time cost per machine —
src/README.mdwalks you through it.) - Django admin is your back-office. Menu items, tables, plans, vendors — managed there, not in hand-built screens.
- No payments, no email/SMS, no user registration. At most one staff login. Customer-facing forms take a name and phone number.
- The AI part is a pure-Python module (
ai/) with plain functions your views call. It must run and be demonstrable on its own, without Django.
The guides are ordered for you in the From zero to your first Python web project roadmap — follow it alongside the lab sessions. Your AI reading depends on your team:
- Cafe, Restaurant, Party Palace → How AI search works
- Gym → How expert systems work (read the search guide first — backward chaining is DFS)