Elliot Dauber
// effect talk
Building LLM Systems with Effect
As a startup building LLM applications at scale, we must balance reliability, user experience, and developer speed. At Markprompt, we’ve built our applications entirely in Effect, using it to solve challenges like LLM hallucination and testing in non-deterministic environments.
This talk covers how we use Effect for stream processing and dependency injection to solve complex problems in our domain. Expect practical insights, real-world examples from our codebase, and a discussion of Effect's role in making our systems more reliable and testable — as well as some unanswered questions.
// bio
Elliot recently graduated from Stanford University, where he earned both a bachelor's and a master's degree in computer science. Before joining Markprompt as a founding engineer, Elliot worked at Humane, focusing on Java and Android development—completely different from his current work with JavaScript and TypeScript.
Outside of tech, Elliot has a passion for music, which he played extensively during school, and loves traveling.