Week 2 of 8 · Build AI

SDK, chat, generative UI and memory

Anatomy of an AI app, from a single API call to a full-stack chat application with generative UI and memory.

What the week builds toward

WorkBot's web chat, running on either provider: streaming responses, tool calls shown in the UI, generative UI from a safe component catalog, per-tenant memory of entities, cases and preferences, deployed on Vercel.

The 5 days

  1. Day 6

    Anatomy of an AI app

    Both SDKs, Anthropic and OpenAI: the sampling controls, streaming, and a CLI chatbot that tracks its own cost.

    You ship: WorkBot's conversation core: a streaming CLI chatbot on either provider, with model switching and cost tracking.

  2. Day 7

    Tool use and chaining calls

    Implement tool use end to end, run multi-turn tool loops, and chain focused LLM calls into a research pipeline.

    You ship: 4 tools running on both providers, and a research command with citations.

  3. Day 8

    Full-stack AI chat app

    Build and deploy a streaming chat app in Next.js, with tool execution visible in the interface.

    You ship: WorkBot's web chat deployed on either provider, with tool status and results in the UI.

  4. Day 9

    Generative UI

    Generate dashboards and forms from a constrained component catalog, rendered as the JSON streams in.

    You ship: An AI dashboard generator with a 15-component catalog and 3 demo scenarios.

  5. Day 10

    Context and memory

    Manage conversation context within token limits, add long-term memory, and ship WorkBot's web chat with memory.

    You ship: WorkBot's web chat with context management and memory, deployed.