About this project

Agents Towards Production is an open-source collection of runnable, code-first tutorials aimed at developers who want to move GenAI agents from prototype to production. Each tutorial lives in its own folder with notebooks or code files, so readers can go from concept to a working agent quickly. The repository describes itself as covering the building blocks of a GenAI agent stack: stateful workflows, vector memory, real-time web search APIs, Docker deployment, FastAPI endpoints, security guardrails, GPU scaling, browser automation, fine-tuning, multi-agent coordination, observability, evaluation and UI development. An architecture diagram in the README illustrates the flow of building a production-level agent, with tutorials mapped to each component. Tutorials are grouped by theme. Tool integration includes secure tool calling with OAuth2 authentication and human-in-the-loop approval workflows. Data processing includes local, private file conversion using a WebAssembly engine plus an MCP tool call for office documents and PDF-to-markdown, and web data collection for agents at scale. Other listed tutorials cover agent memory with Redis and Mem0, RAG with Contextual AI, real-time web search with Tavily, durable RAG ingestion with Inngest, GPU deployment with RunPod, and a Kotlin agent framework tutorial with JetBrains Koog. Several tutorials are contributed by sponsoring companies, and the README links to each sponsor's tutorial. The project also promotes related educational material: a paid course on building software with AI, a free module, a newsletter, YouTube explainer episodes and a Reddit community. These are promotional elements rather than part of the tutorial code itself. Overall, this is a learning resource and reference collection for practitioners working with agent frameworks, memory systems, retrieval, deployment and safety controls. It does not present itself as a library or runtime; its value is in the step-by-step tutorials and the breadth of production concerns it addresses.