Roadmap Map

Main Path

flowchart TD
  S0["0. Orientation<br/>Build the map before choosing tools."]
  S1["1. Foundations<br/>Become fluent with code, data, math, and systems basics."]
  S2["2. Machine Learning<br/>Learn how models learn from data and how evaluation can lie."]
  S3["3. Deep Learning<br/>Train, debug, and reason about neural networks."]
  S4["4. LLMs<br/>Understand tokens, transformers, generation, adaptation, and evaluation."]
  S5["5. AI Applications<br/>Build reliable AI products around models."]
  S6["6. AI Agents<br/>Build controlled tool-using systems, not vague autonomy."]
  S7["7. Model Infrastructure<br/>Operate AI systems as production software."]
  S8["8. Optimization and Hardware<br/>Make inference faster, cheaper, and more predictable."]
  S9["9. Security, Blockchain, ZKML<br/>Secure, govern, and verify AI-enabled systems."]
  S10["10. Mastery<br/>Own a complete AI system end to end."]

  S0 --> S1
  S1 --> S2
  S2 --> S3
  S3 --> S4
  S4 --> S5
  S5 --> S6
  S6 --> S7
  S7 --> S8
  S8 --> S9
  S9 --> S10
  S5 --> S7
  S6 --> S9
  S8 --> S10

Stage Structure

Stage	Part count	Parts
Stage 0: Orientation	5	AI Engineering Mental Model, Tooling and Learning Environment, Use Case Judgment, Learning Operating System, First Portfolio Skeleton
Stage 1: Foundations	6	Python Software Craft, Math for AI Engineers, Data Handling, Databases and Storage, Web and API Basics, Systems Thinking Basics
Stage 2: Machine Learning	5	Problem Framing, Splits and Leakage, Supervised Models, Metrics and Error Analysis, Unsupervised Representations
Stage 3: Deep Learning	6	Neural Network Core, Training Loop Engineering, Optimization Dynamics, Architectures and Modalities, Debugging Neural Systems, Hardware Aware Training Preview
Stage 4: Large Language Models	7	Token and Context Mechanics, Transformer Mental Model, Generation Controls, Model Landscape, Prompting and In-Context Learning, Fine-Tuning and Dataset Engineering, LLM Evaluation Methodology
Stage 5: AI Applications	7	AI Product Interface, Prompt System Engineering, RAG Ingestion, Retrieval and Reranking, Grounded Generation, Application Evaluation and Feedback, Guardrails and Release Architecture
Stage 6: AI Agents	8	Agent Loop Fundamentals, Reasoning and Planning Patterns, Tool Design, MCP and Tool Ecosystems, Memory and Agentic RAG, Multi-Agent Systems, Agent Evaluation and Observability, Agent Security and Safety
Stage 7: Model Infrastructure	6	Data Pipeline Architecture, RAG and Feature Infrastructure, Training and Adaptation Pipelines, Serving and Deployment, Observability and Quality Operations, Reliability and Cost Control
Stage 8: Optimization and Hardware Acceleration	7	Inference Performance Model, Transformer Inference Internals, Model Optimization, Serving Engines, Distributed Inference, GPU and Kernel Basics, Edge and Accelerator Co-Design
Stage 9: AI Security, Blockchain, and ZKML	6	LLM and Agent Security, Secure AI Application Architecture, Governance and Responsible AI, Blockchain Fundamentals, Smart Contract Security, ZK and Verifiable AI
Stage 10: Mastery	5	Capstone Problem and Architecture, Capstone Build Execution, Portfolio Communication, Interview and Collaboration Readiness, Specialization and Research Frontiers