About This Course
<p>This course contains the use of artificial intelligence.</p><p>Every app makes thousands of small decisions. Is this ticket urgent? Which team owns it? Is this message safe to send? Which tool should my AI agent call? For years we answered these with brittle rules or with slow, costly chat models that sometimes return broken JSON.</p><p>In September 2026, TypeSafe AI released Jev, the first System One model. It does not chat and it does not write text. You give it an input and a few typed questions, and it returns typed answers with probabilities in a single fast pass. Within weeks, a whole family of decision models followed, including Laya, an open model that runs on your own laptop.</p><p>This course gives you a clear, beginner-friendly understanding of decision models: what they are, how they work, and where they fit.</p><p><span>What you will learn</span></p><ul><li><p>What a decision model is, who built Jev and why, and how it differs from a chat model</p></li><li><p>The three question types, Choice, Score and Noul, and when to use each one</p></li><li><p>How to read probabilities, confidence and calibration, and how to set thresholds from the cost of a mistake</p></li><li><p>How decision models are trained: RLHF vs RLVR vs RLCD</p></li><li><p>Four production patterns: speculative fan-out, confidence-gated routing, composite scoring and intent routing</p></li><li><p>10 real use cases, from helpdesk triage and prompt injection guards to RAG filtering and payment risk</p></li><li><p>How decision endpoints differ from the chat completions endpoints you already know</p></li><li><p>The landscape of closed and open decision models, and the SDKs, gateways, frameworks and local runtimes that support them</p></li><li><p>Where Jev gets things wrong, and the habits that make decision models reliable</p><p><br></p></li></ul><p><span>Honest by design</span></p><p>Most speed and accuracy numbers you see online come from the companies themselves. This course shows you where decision models shine and where they fail, and teaches you to separate vendor claims from what you can measure yourself.</p><p>No machine learning background is needed, and no code is required. Developers, product managers and founders will all come away with the vocabulary and judgment to plan real features with decision models.</p>
What you'll learn:
- Explain what a decision model is and how Jev, a System One model, differs from a chat model like GPT or Claude
- Choose the right question type, Choice, Score or Noul, for any decision your app needs to make
- Read probabilities, confidence and calibration, and set safe thresholds from your own labelled data
- Use the TypeSafe console, Playground and API to make fast, low-cost typed decisions
- Compare decision models and chat models on 10 real use cases for accuracy, speed and cost
- Run open decision models like Laya locally with Ollaya and Ollama, with no API cost
- Switch the same code across providers such as OpenRouter, Liquid, Perplexity and Cloudflare
- Apply production patterns: fan-out, confidence-gated routing, composite scoring and intent routing