Building Effect-ive Agents | Maxwell Brown (Effect Days 2025)

Get support from the Effect community → https://discord.gg/effect-ts Maxwell Brown – keynote speaker at Effect Days 2025 – discusses building complex agentic systems using Effect and its AI integration packages, emphasizing a practical approach with code examples. He showcases how developers can leverage Effect to build scalable, resilient, and maintainable systems for working with large language models (LLMs). Keynote repo: https://github.com/IMax153/effect-days-2025-keynote/tree/main Anthropic blog post: https://www.anthropic.com/engineering/building-effective-agents 00:00 Intro 02:14 Why Effect for AI systems 03:25 Complexity of production systems 05:00 Non-determinism in LLM-powered apps 06:18 Effect as the antidote 08:00 What is an agentic system? 09:30 Augmented LLMs explained 11:27 Effect AI packages & LLM interactions 15:15 Using multiple models in one app 17:08 Building composable services with AI models 19:00 Providing clients via layers 22:30 Handling failures in AI systems, AI Plan module 26:22 Fine-grained configuration overrides 28:18 Defining and using tool calls 31:32 What’s next for Effect AI & closing thoughts ______________________________________ Maxwell Brown is a Core Contributor at Effect and Founding Engineer of Effectful Technologies. Learn more about and follow Maxwell: X (Twitter): https://twitter.com/imax153 Github: https://github.com/IMax153 ______________________________________ Effect is an ecosystem of tools to build production-grade applications in TypeScript. Website & docs: https://effect.website/ Community & support: https://discord.gg/effect-ts Follow us on X (Twitter): https://twitter.com/EffectTS_ Github: https://github.com/Effect-TS ______________________________________ #Effect #TypeScript #TypeScriptLibrary #AI #LLM