14,800 real, runnable mini-projects — each with working Python code, full documentation, system design (HLD + LLD), a live demo runbook, and a resume/presentation guide. Pick any project, run it, present it, put it on your resume.
Twelve of the strongest projects to try first. Every one runs with zero setup (python3 main.py):
| Project | What it is |
|---|---|
| Decision Tree classifier | Gini-split decision tree built from scratch, trained and evaluated in the demo |
| DBSCAN clustering | Density-based clustering from scratch on generated data |
| Multi-client chat server | Real socket server handling concurrent clients |
| Web dashboard | Live dashboard served by stdlib http.server, no frameworks |
| Encryption-at-rest auditor | CLI that audits files for encryption coverage |
| Monitoring agent | Alert deduplication engine, SRE-style |
| Market-basket pipeline | End-to-end data pipeline running the Apriori algorithm |
| Idempotency key store | LRU-backed idempotency store for APIs |
| SIR epidemic simulator | Epidemiological model with tunable parameters |
| Coupled pendulums | Normal-modes physics simulation |
| BMI trend tracker | Health dashboard with trend analysis |
| Air-sensor calibrator | Linear regression calibrating sensor readings |
A library of 14,800 real, runnable, documented projects written in Python. Every project is a complete working program: it runs with the standard library alone, ships with a seeded demo that reproduces byte-for-byte, and is documented the way a professional team documents software — README, high-level design, low-level design, runbook, and presentation guide. No stubs, no placeholders, no "exercise left to the reader."
| Domain | Projects |
|---|---|
| computer_science | 6,600 |
| biology | 1,000 |
| physics | 800 |
| chemistry | 800 |
| mathematics | 700 |
| healthcare | 1,600 |
| electronics_engineering | 700 |
| environmental_science | 600 |
| ai | 1,200 |
| business_analyst | 800 |
| Total | 14,800 |
The majority are computer science because that is where runnable, dependency-free projects compose best; the remaining seven domains carry the same quality bar — every project runs, every demo is seeded, every document exists.
realtime-projects/
├── README.md # this file
├── LICENSE # MIT
├── COPYRIGHT.md # originality pledge
├── CONTRIBUTING.md # how to add a project
├── docs/
│ ├── REPO_ARCHITECTURE.md # how the repo is organized
│ ├── EXPORT_TO_PDF.md # turn any project's docs into a PDF
│ ├── PRESENTING_GUIDE.md # how to demo a project in 5 minutes
│ └── RESUME_GUIDE.md # how to list these projects on a resume
├── topics/ # machine-readable topic catalog (one JSON per subdomain)
├── templates/docs/ # doc templates the generator fills per project
├── tools/
│ └── generate.py # the project generator (topics × templates → projects)
└── projects/
└── <domain>/
└── <subdomain>/
└── <project-slug>/
├── README.md
├── <entrypoint>.py
└── docs/
├── ARCHITECTURE.md
├── DESIGN.md
├── RUNBOOK.md
└── PRESENTATION.md
# pick any project directory
cd projects/computer_science/algorithms/bubble-sort-visualizer
# the demo runs with zero setup — Python 3.8+, no pip install
python3 main.py
# every flag is documented in the CLI itself
python3 main.py --helpThat is the entire contract: if it is in projects/, it runs.
Each project ships a docs/PRESENTATION.md with a 60-second pitch, architecture talking points, a copy-paste resume bullet, and an interview Q&A. For general guidance on structuring a live demo, read docs/PRESENTING_GUIDE.md.
Any project's full document set (README + all four docs/) can be exported to a single PDF with one command. See docs/EXPORT_TO_PDF.md.
New projects are added through the generator, not by hand: add a topic to topics/<subdomain>.json and run tools/generate.py. Full instructions and the code/docs standards are in CONTRIBUTING.md.
MIT. See LICENSE and the originality pledge in COPYRIGHT.md.