Loading...

AI Engineer Roadmap

You build, train, and deploy AI systems end-to-end

The full AI engineering stack - deep learning, LLMs, agents, and production deployment.

The route

  1. Week 1 - Python for ML: Your primary language - NumPy, pandas, scikit-learn are used daily
  2. Week 2 - AI Tools: Claude Code and OpenClaw multiply every week that follows - which is why they come early, not last
  3. Week 2 - Git & Version Control: Model versioning, experiment tracking, and ML-specific Git workflows
  4. Week 2 - Jupyter & Colab: Rapid prototyping and experiment iteration before production code
  5. Week 2 - SQL for Data: The metadata filter behind a retriever, the source of truth a RAG index was built from, and every text-to-SQL tool you will ship
  6. Week 3 - Foundations: Deep math fundamentals power every design decision you make as an AI engineer
  7. Week 3 - Statistics: Model evaluation, A/B testing, and statistical significance in experiments
  8. Week 4 - Probability Theory: Bayesian reasoning, distributions, and uncertainty quantification in models
  9. Week 4 - Linear Algebra: Embeddings, transformations, PCA - the math behind every neural network
  10. Week 5 - Calculus: Backpropagation, gradients, chain rule - how neural networks actually learn
  11. Week 5 - Optimization: Core skill - gradient descent, learning rates, loss surfaces, convergence
  12. Week 6 - Information Theory: Entropy, KL divergence, cross-entropy - needed before deep learning and LLMs
  13. Week 6 - ML Algorithms: Must-know for interviews - regression, trees, SVMs, ensembles, evaluation
  14. Week 7 - Deep Learning: Your bread and butter - CNNs, attention, transformers, training loops
  15. Week 8 - LLM: Build with LLMs - prompt engineering, RAG, fine-tuning, embedding pipelines
  16. Week 11 - Software Engineering for AI: Evaluation first, then production AI code - testing, architecture, system design, observability and cost
  17. Week 12 - Agentic AI & Multi-Agent Systems: Design agents with tool use, memory, planning and guardrails - and know when a deterministic workflow beats one
  18. Week 15 - MLOps & Deployment: Deploy and scale - Docker, monitoring, CI/CD, model registries
  19. Week 15 - ML Pipeline: Deployment-focused pipeline modules - model serving and monitoring
  20. Week 16 - AI on Cloud: Deploy your GenAI and agentic systems on AWS, GCP, or Azure - Bedrock, Vertex AI, AI Foundry, plus multi-cloud patterns
  21. Week 17 - Interview Q&A: Scenario-based interview prep - production problems that AI companies actually ask
  22. Week 17 - AI System Design: The AI system design round is its own interview, and the one that most often separates senior from mid. Sizing a retrieval index and defending a latency budget is not covered by the LLM or agents tracks