Observability
- Tracing and live telemetry
- Watch for drift in production
- Know what changed, and why
LAUNCHAll 9 courses. Free for a limited time.AI engineering.
GOOD ENGINEERING STARTS WITH PRACTICEAI engineering.
Learn it. Build it.
The big ideas behind AI, made practical.
Hands-on labs, a teacher beside you, and something
real to show for your next two hours.
Sign in to explore all courses. Try AI at Work without signing in ↗
Practice at your paceθ A teacher beside you Build with real models Free during launch
THE PRACTICE LIBRARY
Big ideas. Small, working things.
Pick a pattern. Open a lab.
Leave with something you built and understand.
9 labs ready when you areSELF-PACED · HANDS-ON · AI-GUIDED
Harness
- The agent loop, from scratch
- Tools, history, and memory
- Guardrails that hold up
LAB 01·~2 HOURS·TA: Loop·SIGN-IN REQUIRED
Evals
- Rubrics and LLM judges
- Find the failures that matter
- Turn quality into a CI gate
LAB 02·~2 HOURS·TA: Cal·SIGN-IN REQUIRED
Context
- Embeddings to hybrid search
- Chunking and retrieval
- A working RAG pipeline
LAB 03·~2 HOURS·TA: Verba·SIGN-IN REQUIRED
MCP
- JSON-RPC, inside and out
- Tools, resources, and prompts
- The protocol in production
LAB 04·~2 HOURS·TA: Wire·SIGN-IN REQUIRED
Fine-tune
- When fine-tuning makes sense
- Train a real LoRA adapter
- Measure it against the base
LAB 05·~2 HOURS·TA: Forge·SIGN-IN REQUIRED
Claude Code
- Plans, skills, and subagents
- Work with a real codebase
- Ship a reviewed pull request
LAB 06·~2 HOURS·TA: Pair·SIGN-IN REQUIRED
Voice
- The whole real-time stack
- Latency, VAD, and barge-in
- Take the agent to the phone
LAB 07·~2 HOURS·TA: Echo·SIGN-IN REQUIRED
Generative UI
- From intents to components
- Streaming and hydration
- Interfaces shaped by agents
LAB 08·~2 HOURS·TA: Canvas·SIGN-IN REQUIRED
AI at Work
- No coding required
- Eight useful work artifacts
- A fresh final case and portfolio
LAB 09·~4–6 HOURS·TA: Clara·FREE · NO SIGN-IN
The library keeps growing.Up next: ObservabilityIN THE WORKS
02 THE WAY WE LEARN HEREUnderstand it.
Understand it.
Make it yours.
An illustrated field guide. A place to practice.
And a teacher right where the question comes up.
A short warm-up. A drawing you can follow. Get a feel for the idea before you open the lab.
Small steps.
Something real to show for them.
theta./Generative UISTUDY ROOM
ILLUSTRATED FIELD GUIDEEXAMPLE
An answer you can interact with.
The model chooses a component from a trusted catalog. Your application validates it and brings it to life.
THE IDEA TO TAKE WITH YOU
The model picks from a menu. It doesn’t paint the plates.
01 / 04ONE IDEA. YOUR OWN PRACTICE.
A NOTE FROM THETA
One connected idea. A little clearer with every chapter.
YOUR NEXT USEFUL STEP
Start with where you are.
Use AI in your work, or learn to build with it.
Follow the foundations, then choose your next specialty.
AI AT WORK · FREE · NO SIGN-IN NEEDEDMake a better working day.
- 09AI at Work
- Research, writing, meetings, data, and a portfolio you keep.
02 — CHOOSE A SPECIALTY · SIGN-IN REQUIREDPick the depth your next project needs.
Your progress saves in this browser.Open your workspace
A FEW GOOD QUESTIONSBefore you
Before you
open a lab.
A little context before
you get your hands on the code.
Theta is your AI teacher across the whole curriculum. In machine learning, θ is the symbol for the parameters being learned during training — so we named the teacher after the thing being shaped. Each lab also has its own AI TA (Loop for agents, Cal for evaluation, Verba for context, Wire for protocols). They're focused on their domain; Theta keeps the through-line.
Engineering labs are designed around a focused practice session. AI at Work is a broader course: allow roughly 4–6 hours across its field guide, eight practice artifacts, and final case. These are planning estimates; your work saves in this browser so you can return later.
AI at Work requires no coding and includes a fictional source packet, drafting tools, and exportable artifacts. The engineering labs assume familiarity with programming; they focus on agent loops, evaluation, retrieval, and integrations.
The courses combine illustrated explanations, practice, and contextual AI help. AI at Work adds one connected fictional project, source-based fact checks, editable work artifacts, and a different final case. Judge the learning by the work you can produce and explain; automated practice feedback is not a professional accreditation.
No. The artifact is the certificate — you finish each lab having built a working thing you understood. If your employer wants proof, share the lab URL plus what you built.
AI at Work is free to explore without signing in. All other courses require a Theta account. Payments and paid-course access will be introduced separately; signing in does not purchase a course.
Observability is next in the library. New labs keep the same format: a focused pattern, a working artifact, and an AI TA to help you understand what you built.
An independent engineering studio — Theta Studio. The curriculum is anchored on canonical sources (Karpathy, Anthropic, Ng, Husain, Willison) and built with the discipline you'd want in production systems, not in vibe-y demos.