AI Engineer
Ship LLM features, not just prompts.
The complete applied-AI path: what the role actually is, the model landscape, working with model APIs and tokens, prompt engineering, safety, open-source models, embeddings and vector databases, RAG, agents, multimodal AI and the modern AI dev toolchain.
Saved on this device — no account needed.
- 1
Introduction
Know the role, the vocabulary and the landscape.
1 week0/5 - 2
Pre-trained Models
Choose models like an engineer, not a fan.
1–2 weeks0/4 - 3
Working with Model APIs
Call models reliably, safely, affordably.
2–3 weeks0/6 - 4
Prompt Engineering
Treat prompts as versioned artifacts, not vibes.
2 weeks0/4Build: Structured extraction service
An endpoint that turns messy text (resumes, invoices) into validated JSON with retries and cost logging.
- 5
AI Safety & Ethics
Ship features you can defend to security review.
1–2 weeks0/5 - 6
Open-Source AI
Run models you control.
2 weeks0/4 - 7
Embeddings & Vector Databases
Give models a searchable memory.
2–3 weeks0/5 - 8
RAG
Ground models in data they weren't trained on.
2–3 weeks0/4Build: Docs chatbot with citations
RAG over a real documentation set; every answer cites its sources and admits when nothing relevant exists.
- 9
AI Agents
Automate multi-step work with control you can defend.
2–3 weeks0/5 - 10
Multimodal AI
Beyond text: images, audio and speech.
1–2 weeks0/4 - 11
Evals & Production
Prove quality; run it live.
2 weeks0/4 - 12
Ship & Prove
A deployed AI product and verified proof of skill.
2 weeks0/2