17-Module Course

AI Engineer On-Demand

Build production AI systems in TypeScript — no Python required. From transformers to deployed agents, RAG pipelines, and multimodal generation.

Modules

~18h

Est. Time

Intermediate

Skill Level

0 / 17

Completed

Course Progress 0%

Start Learning →

Part 1 — LLM Foundations

Modules 1–3

Part 1: Foundations

Module 01

How Transformers Work

Trace data through the architecture you use every day — from raw text to predicted token.

Part 1: Foundations

Module 02

LLM Controls

Master the parameters that shape every response — and develop the intuition to set them deliberately.

Part 1: Foundations

Module 03

Training and Post-Training

Understand how Claude got its capabilities — and when you should (and shouldn't) change them.

Part 2 — RAG and Prompting

Modules 4–5

Part 2: RAG

Module 04

RAG Systems

Give your LLM the right context at the right time — build a production-quality retrieval pipeline.

Part 2: Prompting

Module 05

Prompt Engineering

Move from ad-hoc prompting to a systematic, tested approach that produces consistent results.

Part 3 — Agents and Tools

Modules 6–9

Part 3: Agents

Module 06

Agent Fundamentals

An agent is just an LLM in a loop with tools — understand the loop before you build one.

Part 3: Tools

Module 07

Tool Calling

Build tools Claude can use — master the API pattern that powers every agentic system.

Part 3: Patterns

Module 08

Agentic Patterns

Five patterns from Anthropic's framework — learn to recognize them and choose the right one.

Part 3: MCP

Module 09

MCP — Model Context Protocol

The open standard that lets any tool connect to any model — build your own MCP server.

Part 4 — Reasoning and Research

Modules 10–12

Part 4: Reasoning

Module 10

Reasoning Models

Extended thinking changes what's possible — learn when to use it and how to configure it correctly.

Part 4: Scaling

Module 11

Inference-Time Scaling

Trade compute for quality selectively — know which technique to reach for and when to stop.

Part 4: Research

Module 12

Deep Research Systems

Build a system that researches 10 sources in parallel and synthesizes them into a structured report.

Part 5 — Multimodal

Modules 13–15

Part 5: Diffusion

Module 13

Generative Model Landscape

Understand the four families of generative models — and why diffusion currently wins.

Part 5: Image APIs

Module 14

Image Generation APIs

Go from prompt to deployed image generation feature in under 3 hours using Replicate and Flux.

Part 5: Vision

Module 15

Multimodal Integration

Close the loop — use Claude Vision to analyze what you generate and refine it automatically.

Part 6 — Build and Capstone

Modules 16–17

Part 6: Capstone

Module 16

Capstone Project

Design and build your own AI system from scratch — no scaffolding, just your architecture decisions.

Part 6: Reflection

Module 17

Reflection and Next Steps

Measure what you learned, identify your next frontier, and plan how to keep building.