AI Engineer Study Library

6-month plan: from the ground up

Twenty-four weeks for people starting with little programming: a month of software basics (Python, Git, SQL and APIs), then the 3-month plan's capstone, an assistant over documents you know, with extra weeks for a web front end, data pipelines, production RAG, agent frameworks, Docker, the cloud, Kubernetes, system design and forward-deployed work. About 8 hours a week. The topics follow the learning path; other plans: 1-month plan · 2-month plan · 3-month plan.

An unofficial study plan built from the public posts of Bashiri Smith and Melvin Vivas, not affiliated with or endorsed by them. Items under Fill the gap are not from the creators: they cover skills their posts don't. Track your progress in the library app; ticks are saved in your browser.

Week 1: Map the role and plan

Goal Explain what AI engineers build and get hired for, rate your skills, and write a capstone brief with 20 real test questions. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Week 2: Python for AI work

Goal Write and run inventory.py in its own virtual environment, using functions, lists and dictionaries, file reading, try/except and a small class, with its dependencies in requirements.txt. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 3: Git, GitHub and SQL

Goal Keep your work in Git and on GitHub with branches and pull requests, and store and query data in SQL, joins included. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 4: APIs and a Python backend

Goal Explain how web APIs work, call one from Python, and serve your own data through a typed FastAPI service. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 5: How neural networks learn

Goal Explain how a neural network learns (loss, gradient descent, backpropagation) and train a tiny network you wrote from scratch. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 6: LLM fundamentals and APIs

Goal Explain tokens, training and context windows, call hosted and local LLMs from Python, and choose a model on your own questions. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 7: Prompts and coding agents

Goal Write testable prompts that return validated JSON, and use a coding agent with an AGENTS.md and plan mode. (about 9 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 8: A web front end for AI apps

Goal Build a small React page that talks to your API, so the capstone has a face a hiring manager can click. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 9: Data pipelines

Goal Load data from files and APIs incrementally, so re-runs only process what changed, and explain where an event stream such as Kafka fits. (about 8 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 10: Embeddings and vector search

Goal Embed, store and search text in a vector database, explain HNSW, and pick an embedding model by measuring retrieval on your data. (about 8 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Week 11: RAG that holds up

Goal Build RAG that answers only from your documents with citations, and improve it with chunking, hybrid search and reranking. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 12: Production RAG and RAG evaluation

Goal Measure a RAG pipeline like production code: score it with faithfulness, answer relevancy and contextual recall, find the three worst answers, and fix one cause with before-and-after numbers. (about 7 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 13: Agents, tools and MCP

Goal Build a bounded tool-using agent, serve a tool from your own MCP server, and pick the simplest agent pattern that works. (about 9 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 14: Agent frameworks: LangGraph and MCP servers

Goal Rebuild your agent as a LangGraph state graph that calls your week-13 MCP tools, and judge from steps, cost and failures when a framework beats plain code. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional, to fill the gap (not from the creators)

Week 15: Evals and guardrails

Goal Measure retrieval and answers with a golden set and a checked LLM judge, and defend against prompt injection, unsafe input and PII leaks. (about 9 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 16: Docker and CI/CD

Goal Package the capstone in containers and let CI test and build it on every push. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 17: Ship, trace and control cost

Goal Deploy a containerized API with tracing, a fallback model, caching and a CI eval gate; report P50/P95 latency and cost per query. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 18: Cloud fundamentals and infrastructure as code

Goal Deploy the container to a cloud provider safely, with secrets kept out of the image and every resource described in code. (about 8 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 19: Kubernetes and monitoring

Goal Run the app on Kubernetes with health checks and replicas, and watch latency and cost with Prometheus and Grafana. (about 8 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Week 20: Fine-tuning and system design

Goal Decide when to fine-tune, fine-tune a small router model and compare it with an LLM, and present your design trade-offs. (about 8.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional

Optional, to fill the gap (not from the creators)

Week 21: System design: services, auth and scale

Goal Design the capstone for a whole company: services, sign-in and per-user permissions, queues, caching and failure handling. (about 7.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional, to fill the gap (not from the creators)

Week 22: Forward-deployed engineering and talking to users

Goal Explain the forward-deployed engineer role, run user interviews that reveal real problems, and pitch what you built in three minutes. (about 6.5 h in total)

Core

From the creators' posts

Build

Fill the gap (not from the creators)

Optional, to fill the gap (not from the creators)

Week 23: Portfolio week

Goal Show a deployed, documented and evaluated project with a demo and a case study a hiring manager can scan in two minutes. (about 7.5 h in total)

Core

From the creators' posts

Build

Optional

Optional, to fill the gap (not from the creators)

Week 24: Job search and trends

Goal Run a targeted job search, answer interview questions with numbers from your project, and explain where AI engineering roles are heading. (about 8 h in total)

Core

From the creators' posts

Build

Optional

Optional, to fill the gap (not from the creators)