Skip to content

An archive of all my learnings to become an applied ML systems engineer

Notifications You must be signed in to change notification settings

Bantuson/ai-engineer-roadmap

Repository files navigation

AI Engineer Learning Roadmap

A comprehensive learning repository for AI engineering, agent development, product management, and interview preparation for technical teams.

What's Inside

1. Agent Frameworks (agent-frameworks/)

Documentation and hands-on mini-projects for 10 AI agent frameworks:

  • CrewAI, LangGraph, LangChain, OpenAI SDK, Google ADK
  • Claude SDK, LlamaIndex, AutoGen, Semantic Kernel, Pydantic AI

Each mini-project demonstrates a different framework pattern—from multi-agent systems to RAG implementations.

2. Product Management (product-management/)

A comprehensive, self-study product management course for technical teams:

  • Foundations — Core PM concepts (user research, strategy, roadmapping, metrics)
  • Technical PM — Working with engineering teams, writing PRDs, APIs
  • AI Product Management — Managing AI/ML products, evaluation, responsible AI
  • Prompt Engineering for PMs — Using AI for productivity, production prompts
  • Frameworks & Tools — Prioritization, stakeholder management, modern PM toolkit
  • Capstone Projects — Portfolio-ready projects demonstrating PM skills

Designed for developers transitioning to PM or technical professionals working alongside product teams.

3. Agent Evals (evals/)

Production-grade LLM evaluation course with practical scripts:

  • Evaluation Fundamentals — Why evals matter, metrics, eval types
  • Eval Frameworks — RAGAS, DeepEval, custom scripts, LLM-as-judge
  • Testing AI Features — Unit tests, integration tests, golden datasets
  • A/B Testing & Monitoring — Production monitoring, drift detection
  • Eval Scripts — Ready-to-run evaluation scripts for all mini-projects

Includes shared modules for metrics (BLEU, ROUGE, F1), LLM-as-judge, and reporting.

4. AI Security (vulnerabilities/)

AI security course covering threats, attacks, and defenses:

  • Threat Landscape — OWASP Top 10 for LLMs, attack surface mapping
  • Prompt Injection — Direct/indirect attacks, defense strategies
  • Jailbreaking — Techniques, guardrail bypasses, defense mechanisms
  • Data & Model Security — Data poisoning, model extraction, adversarial attacks
  • Defense Strategies — Input/output validation, rate limiting, monitoring
  • AI Red Teaming — Methodology, attack playbooks, capstone exercise
  • Attack Library — Reference patterns for security testing

5. Interview Prep (behavioral/)

Comprehensive interview preparation for AI Engineer and AI PM roles:

  • Behavioral Interviews — STAR method, leadership principles, storytelling
  • DSA/LeetCode — Pattern-based approach (50+ problems across 9 categories)
  • AI Engineer Interviews — Python proficiency, frameworks, prompt engineering, system design
  • AI PM Interviews — Product sense, technical depth, AI-specific challenges

Includes question banks, mock interview guides, and mental models.

6. Learning Workflow (learning-workflow/)

3-week intensive study program (Jan 19 - Feb 7, 2026):

  • Daily Schedule — 13-hour weekday structure, 6-hour weekend plans
  • Week-by-Week Plans — Daily objectives, exercises, deliverables
  • Prompt Drills — 240 prompts across agent design, security, evaluation
  • Trackers — Progress tracking, skill checklists, deliverables tracker

Structured to build Python, PM, DSA, and AI engineering skills in parallel.

7. Agent Roles Library (agent-roles-library/)

Library of reusable agent role definitions in XML/Markdown format for use with AI systems.

8. AI Agent Design Patterns (ai-agent-design-patterns/)

Documentation of common AI agent patterns: ReAct, RAG, multi-agent, tool use, and more.

9. Python Foundations (python/30_day_foundation/)

Python fundamentals exercises organized as a 30-day learning path.

Quick Start

Running AI Agent Mini-Projects

# Set API key
export DEEPSEEK_API_KEY="your-key"

# Install dependencies
cd agent-frameworks/mini-projects/
pip install -r requirements.txt

# Run a project
cd 09_pydantic_ai_calculator
python main.py

Starting the PM Course

Begin with product-management/README.md for the learning path overview.

Repository Structure

ai-engineer-roadmap/
├── agent-frameworks/
│   ├── documentation/           # Framework documentation
│   └── mini-projects/           # 10 hands-on projects
├── product-management/
│   ├── 01-foundations/          # PM fundamentals
│   ├── 02-technical-pm/         # Working with engineering
│   ├── 03-ai-product-management/# AI/ML product skills
│   ├── 04-prompt-engineering-pm/# AI for PM workflows
│   ├── 05-frameworks-tools/     # PM frameworks and toolkit
│   ├── 06-capstone-projects/    # Portfolio projects
│   └── resources/               # Templates and reading list
├── evals/
│   ├── 01-evaluation-fundamentals/  # Core eval concepts
│   ├── 02-eval-frameworks/          # RAGAS, DeepEval, LLM-as-judge
│   ├── 03-testing-ai-features/      # Unit/integration tests
│   ├── 04-ab-testing-monitoring/    # Production monitoring
│   ├── 05-capstone-eval-system/     # Capstone project
│   └── scripts/                     # Runnable eval scripts
├── vulnerabilities/
│   ├── 01-threat-landscape/         # OWASP Top 10, attack surface
│   ├── 02-prompt-injection/         # Injection attacks & defenses
│   ├── 03-jailbreaking/             # Jailbreak techniques
│   ├── 04-data-model-security/      # Data/model attacks
│   ├── 05-defense-strategies/       # Defense patterns
│   ├── 06-ai-red-teaming/           # Red team methodology
│   └── attack-library/              # Attack reference patterns
├── behavioral/
│   ├── 01-behavioral-interviews/    # STAR method, leadership
│   ├── 02-leetcode-dsa/             # Pattern-based DSA prep
│   ├── 03-ai-engineer-interviews/   # Technical AI interviews
│   └── 04-ai-pm-interviews/         # AI PM interviews
├── learning-workflow/
│   ├── overview/                    # 3-week roadmap, schedules
│   ├── week1/                       # Jan 19-25 daily plans
│   ├── week2/                       # Jan 26-Feb 1 daily plans
│   ├── week3/                       # Feb 2-7 daily plans
│   ├── prompt-drills/               # Daily prompt exercises
│   └── trackers/                    # Progress tracking
├── agent-roles-library/         # Reusable agent definitions
├── ai-agent-design-patterns/    # Design pattern documentation
└── python/                      # Python learning exercises

Contributing

Contributions are welcome! Please see individual folder READMEs for specific guidelines.

License

This educational content is provided for learning purposes.

About

An archive of all my learnings to become an applied ML systems engineer

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors