Learn the modern data + AI stack by doing — six interactive courses that run entirely in your browser.
No signup. No install. No video. Open index.html and start learning — many lessons embed live, in-browser simulators you can poke, break, and rebuild.
One of 40+ live, in-browser simulators — here, a Kafka consumer group rebalancing in real time. No video, no setup: you drive it.
- Zero signup, zero install — no account, no email wall, no
npm install. Clone or click and learn. - Runs 100% in the browser — every course is static HTML/CSS/JS. No backend, no API keys, no database.
- Real interactive simulators, not videos — change an input, watch the system respond: Kafka consumer groups, event-time watermarks, CDC flows, partition skew, bloom filters, A/B-test power, causal DAGs, and more.
- Works offline — open
index.htmlfrom disk with no network and it still works. - Open source, MIT — fork it, remix it, teach from it.
Bloom filter — O(1) membership; bits light up as you add keys (Data Infrastructure) |
CAP, live — cut the cable, watch replicas diverge (Data Infrastructure) |
Bias–variance — slide complexity, watch overfitting fan out (Data Science) |
These aren't videos — open any course and drive them yourself.
Data Engineering Fundamentals 🏗️ flagshipBuild production-grade data pipelines from the ground up: ETL patterns, batch + streaming, partitioning, orchestration, and data quality. 10 chapters · ~15 live simulators
|
From exploratory analysis to model deployment: statistical thinking, CLT, bias/variance, ROC/PR, SHAP, A/B-test power, causal DAGs, drift — plus a full capstone. 13 chapters · live sims throughout
|
The data stack at staff-engineer system-design depth: storage internals, CAP/PACELC, modeling, Parquet/ORC/Avro, lakehouse (Iceberg/Delta/Hudi), streaming + watermarks, CDC/Lambda/Kappa, idempotency, SLAs. 12 lessons · 7 interactive widgets
|
A terminal-first playbook for the OpenAI coding agent. Mental model, sandboxing, 12 lessons + capstone
|
Prompt like you mean it — the handful of habits that separate people who love Claude from people who bounce off it: prompt anatomy, context engineering, 12 lessons
|
An operating model for working AI-native: the mindset, engineering practice, and org design for shipping with AI agents. A hash-routed journey with interactive exercises and quizzes. 9 modules · 39 lessons
|
There is no build step. No bundler, no transpiler, no node_modules. Clone and open a file.
git clone https://github.com/Mavengence/interactive-courses.git
cd interactive-courses
# Option A — just open the landing page in your browser
open index.html # macOS (use `xdg-open` on Linux, `start` on Windows)
# Option B — serve it (recommended; keeps relative paths happy)
python3 -m http.server 8080
# then visit http://localhost:8080/Each course lives in its own folder and ships its own index.html, so you can jump straight in:
open data-engineering-fundamentals/index.html
open data-science/index.html
open data-infrastructure/index.html
open codex/index.html
open claude/index.html
open ai-native/index.html- You learn by doing. Live, in-browser simulators turn abstract ideas — watermarks, CAP trade-offs, partition skew, statistical power — into things you can manipulate and feel.
- Staff-level depth, not a tutorial. The infrastructure and data-engineering courses go to system-design depth used in senior interviews and real platform work.
- Genuinely offline. No CDNs to break, no telemetry, no login. The whole thing is yours on disk.
- Open source. MIT licensed — read the code, fork a course, adapt it for your team, or contribute back.
Improvements of every size are welcome — a typo fix, a clearer explanation, a better simulator, or a whole new lesson. Start with CONTRIBUTING.md, then open an issue or send a PR.
If a simulator made something finally click, leave a star — it takes a second and it genuinely helps other engineers find these courses.
MIT — free to use, adapt, and share. Built by Tim Löhr · @Mavengence.