A Noob Dev productSri Lanka’s fastest-growing EdTech company

Don’t just watch AI/ML tutorials.
Learn it by doing.

Modelplay is a hands-on playground where you write real machine-learning code, run it instantly, and see the model train in front of you. No setup. No servers. Walk away with a portfolio that proves you can build.

epoch0/40loss0.69accuracy52%

classification — a boundary forming between two classes

YOU’LL TRAIN WITH
The loop

Four steps. One that you actually feel: the moment it learns.

  1. 01

    Write

    Fill the blanks in real scikit-learn code. No blank page, no setup paralysis — just the line that matters.

  2. 02

    Run

    Hit run. Python executes in your browser through WebAssembly. No installs, no servers, nothing to break.

  3. 03

    See

    Watch the model train and the plot render in front of you. Instant, visual proof that it worked.

  4. 04

    Prove

    Every pass becomes a portfolio artifact — code, result, and metric — you can send straight to an employer.

modelplay-intro · 90s
A 90-second tour of Modelplay
Real datasetsHidden tests grade your workWorks offline after first loadA portfolio that proves it
Get started

Start free. Run your first model today.

Create your account and jump straight into the labs — no setup, no card. When you’re ready for the full 30-day roadmap, every challenge, and a shareable portfolio, start a 7-day Pro trial.

Free plan, forever · 7-day Pro trial · No credit card to start.

The stack you’ll actually master
Python
pandas
scikit-learn
PyTorch
TensorFlow
Jupyter
Hugging Face
OpenAI
The challenges

7 days of doing, not watching.

Pick a track and commit to one build a day for a week. Real datasets, real code, and a finished project to show at the end — free to start.

7 Days of ML

7 days · live
Start · Day 1 →
1234567

Load data, train models, and ship a real classifier — one build a day.

7 Days of AI

7 days
Preview the track →
inputoutput

From prompts to a working AI feature you can actually demo by day 7.

7 Days of MLOps

7 days
Preview the track →
retraintrainpackagedeploymonitor

Package, deploy, and monitor a model end to end, step by step.

Career roles

Train for the job, not just the topic.

Pick a discipline and climb the ladder employers actually hire for — Intern to Senior — with role-realistic task labs at every tier. Launching soon.

Machine Learning

Coming soon

Climb the ML ladder employers hire for — from your first supervised reps to owning models in production.

  1. 1Intern Machine Learning EngineerClean data, train your first models, and read a metric like a pro.Intern
  2. 2Associate Machine Learning EngineerOwn a model end to end — features, validation, and honest evaluation.Associate
  3. 3Machine Learning EngineerShip models that hold up — tuning, pipelines, and real trade-offs.Engineer
  4. 4Senior Machine Learning EngineerLead the hard calls — architecture, scale, and production reliability.Senior
Preview the Machine Learning ladder →

Artificial Intelligence

Coming soon

Build with real models — from your first prompt to shipping agentic AI features that ship.

  1. 1Intern AI EngineerPrompt, parse, and wire a model into a working first feature.Intern
  2. 2Associate AI EngineerStructured outputs, retrieval, and evals you can actually trust.Associate
  3. 3AI EngineerTools, agents, and RAG systems that do real work reliably.Engineer
  4. 4Senior AI EngineerOwn the AI stack — cost, latency, safety, and architecture at scale.Senior
Preview the Artificial Intelligence ladder →
Built by Noob Dev experts

Learn from people who’ve built it.

Practitioners and teachers behind Modelplay — the people writing the labs, designing the challenges, and answering when you get stuck.

Kavindu Sandaruwan

Kavindu Sandaruwan

Founder & CEO

Founder of Noob Dev, on a mission to make machine learning something you learn by building, not just watching. He leads Modelplay's vision — hands-on labs, real datasets, and a portfolio that proves you can ship — built for Sri Lanka's next generation of AI engineers.

Certified by Noob Dev

Finish a track. Earn a badge that proves it.

Complete a 7-day challenge and walk away with a verifiable Certified Practitioner badge — plus a portfolio of real projects in the tools employers actually use.

+ more
Secure your place →
Machine Learning — Certified Practitioner badgeArtificial Intelligence — Certified Practitioner badge