AI Engineering Roadmap

A comprehensive roadmap outlining the key skills needed to become a successful AI engineer, focusing on practical application and industry best practices. This roadmap incorporates principles of spaced repetition, interleaving, and feedback mechanisms to foster long-term retention and skill mastery. It addresses potential psychological barriers to learning and incorporates social learning aspects for enhanced motivation and engagement.

Powered byDeedit Logo

Program Modules

🤖

Working with LLMs

Understand different LLM providers (OpenAI, Anthropic, Meta, Google), their APIs, and various model types. Learn about streaming, batch processing, and prompt caching. Explore running your own models using platforms like OpenRouter or Ollama. This activity incorporates spaced repetition and interleaving techniques for improved learning.

Understanding LLM Personalities

Daily

Learn how different LLMs have different strengths (e.g., OpenAI for analysis, Anthropic for writing, Gemini for search).

We found that the OpenAI models tend to be the best analysts, the Anthropic models tend to be the best writers, and the Gemini models broadly tend to be the best detectives.

activity

Working with LLM APIs

Daily

Familiarize yourself with common API calls (e.g., openai.Completion.create) and concepts like streaming, batch processing, prompt caching.

activity

Local vs. Open Source Models

Weekly

Explore running your own models using platforms like OpenRouter or Ollama.

activity
✍️

Prompt Engineering

Master the art of prompting to elicit desired behaviors from LLMs. Learn techniques like Chain of Thought, providing examples, and using structured outputs. This activity includes examples and structured exercises to reinforce learning.

Chain of Thought Prompting

Daily

Practice prompting LLMs to explain their reasoning process before providing an answer.

activity

Example-Based Prompting

Daily

Learn how to use examples to guide the LLM's response.

activity

Structured Output Prompting

Daily

Practice prompting for structured outputs (JSON, tables) for easier data processing.

activity
🔎

Retrieval Augmented Generation (RAG)

Learn to combine LLMs with external knowledge sources for enhanced context awareness. This includes activities on embeddings and semantic search.

Retrieval Augmented Generation (RAG)

Weekly

Learn to combine LLMs with external knowledge sources for enhanced context awareness. This includes activities on embeddings and semantic search.

activity
⚙️

LLM Orchestration

Learn to build systems that combine multiple LLMs and tools. Explore frameworks like LangChain and the concept of Agents.

LLM Orchestration

Weekly

Learn to build systems that combine multiple LLMs and tools. Explore frameworks like LangChain and the concept of Agents.

activity
📊

Evaluation and Observability

Develop robust evaluation and monitoring strategies for LLM applications. This includes cost management and performance tracking.

Evaluation and Observability

Weekly

Develop robust evaluation and monitoring strategies for LLM applications. This includes cost management and performance tracking.

activity
🧠

AI Engineering Mindset

Develop a mindset for building with AI, focusing on iterative development, rapid prototyping, and understanding the evolving tool stack. Includes strategies for overcoming procrastination and building self-compassion.

Build First, Build Quickly

Daily

Embrace the iterative process and prioritize rapid prototyping.

activity

Overcoming Procrastination

Weekly

Implement techniques to overcome procrastination and maintain momentum.

activity

Self-Compassion Exercises

Weekly

Practice self-compassion to manage anxiety and build resilience.

activity