INTERMEDIATE
From your first API call to production-grade AI features — authentication, model selection, tool use, function calling, prompt caching, embeddings, and the patterns that separate prototypes from products.
Authentication, request structure, model selection, and the core parameters that control LLM behaviour — the foundation every LLM API developer needs.
Give your LLM the ability to act — define tools, handle multi-turn tool loops, run parallel tool calls, and integrate external APIs into your AI application.
Turn text into meaning — embedding models, cosine similarity, vector database integration, and building a semantic search layer for your application.
Ship AI features that survive the real world — streaming responses, prompt caching, rate limit handling, cost optimisation, and structured output parsing.
A full-stack capstone challenge — design and implement FlowAssist, a document-aware AI assistant with tool use, caching, and streaming, then evaluate its production readiness.