Master AI Engineering.
Get Hired.
The only practice platform built for ML engineers. Solve real problems, get Claude-powered code review, and prove your skills — across 14 AI domains.
Used by engineers at
Why Coderai
Built for the AI engineering interview
LeetCode prepares you for data structures. Coderai prepares you for ML systems.
AI Code Review
Claude analyses every submission — correctness, complexity, best practices — and returns a 0–100 score with specific improvement suggestions.
Real Code Execution
Your Python runs against hidden test cases in a sandboxed environment. See runtime, memory usage, and exactly which cases fail.
14 AI Domains
From NLP and RAG to MLOps and RL — problems are curated from real ML engineering interviews at AI-first companies.
Gamification
XP multipliers, daily streaks, five leagues (Bronze → Diamond), and 20+ achievements keep you coming back every day.
GitHub Intelligence
Link your GitHub and get a 0–1000 developer score based on your activity, open source contributions, AI domain focus, and code quality.
Learning Paths
Structured problem sequences — RAG Engineer, NLP Foundations, MLOps Track — so you always know what to solve next.
How it works
From zero to hired in three steps
Pick a problem
Browse 700+ problems across 14 AI domains. Filter by difficulty, domain, or topic. Or start with a curated learning path.
Write your solution
Code in the browser with Monaco Editor. Hit ⌘↵ to run against sample cases, ⌘⇧↵ to submit for full evaluation.
Get AI feedback
Claude reviews your accepted code instantly — scoring it 0–100 and telling you exactly what to improve. XP hits your account. Streak continues.
14 Domains
Every corner of AI engineering
Generalist or specialist — we have problems that match your role and your next interview.
NLP & LLMs
84 problems
Machine Learning
120 problems
RAG Systems
47 problems
MLOps & Infra
63 problems
AI Agents
38 problems
Math for AI
55 problems
Python & NumPy
72 problems
Fine-Tuning
29 problems
Evals
34 problems
System Design
41 problems
SQL & Data Eng
58 problems
Computer Vision
45 problems
Reinforcement Learning
22 problems
Statistics
61 problems
Leaderboard
Compete with the best
Weekly leagues reset every Monday. Climb from Bronze to Diamond and earn your place on the global board.
alex_ml
312 solved
llm_ninja
287 solved
rag_builder
264 solved
pytorch_pro
241 solved
mlops_dev
228 solved
Pricing
Start free. Upgrade when ready.
No credit card required to start. Cancel Pro anytime.
Free
Everything you need to get started.
- 10 submissions per day
- All 14 domains
- Basic AI hints (3 per problem)
- Leaderboard access
- Streaks & XP
- Community discussions
Pro
For engineers serious about levelling up.
- Unlimited submissions
- Full AI code review (0–100 score)
- All learning paths
- GitHub Intelligence score
- AI hint generation (unlimited)
- Priority queue for execution
- Profile badge
Teams
For engineering teams and bootcamps.
- Up to 10 seats (€9/seat after)
- Everything in Pro
- Admin dashboard
- Custom learning paths
- Team leaderboard
- Usage analytics
- Priority support
FAQ
Questions? Answered.
Coderai is a practice platform for AI engineers. You solve real ML, LLM, and systems problems, run your code against test cases, and get instant AI-powered code review from Claude. Think LeetCode — but built specifically for AI/ML roles.
Yes. The free tier gives you access to all domains with 10 submissions per day, basic hints, and the leaderboard. No credit card required.
Python is the primary language — it covers 95% of real ML/AI interviews. Support for Julia and C++ is on the roadmap.
After each accepted submission, Claude analyses your code for correctness, readability, time/space complexity, and best practices. It returns a score from 0–100 and specific, actionable suggestions — not generic tips.
Your 0–1000 score combines five weighted components: your platform activity (35%), GitHub contribution activity (20%), open source contributions (20%), AI domain relevance of your repos (15%), and code quality based on AI review scores (10%).
Absolutely. Problems are curated from real interview rounds at AI-first companies. Domains include system design, ML theory, coding, statistics, and more — covering the full loop.
Your streak resets to zero. We'll send you an optional reminder email before midnight if you're at risk. You can disable this in Settings → Notifications.
You start in Bronze. Each week, your XP is tallied and compared against others in your league. Top performers get promoted; the bottom group gets relegated. Leagues reset weekly — Diamond is the highest tier.
Ready to level up?
Join thousands of AI engineers practising every day. Start free, no card required.
Start Practising Free