Luma is looking for an engineer to lead our AI Agent workstream. Luma’s future product uses Agents as a Creative Partner, to help our users in exploring ideas for imagination and curiosity across a wide range of topics and interests . We need a strong generalist engineer who can take charge of the agent workstream.
You will primarily work with LLMs and Visual models to expand the capabilities like tool usage/reasoning, real-time world knowledge, understanding of intent and long term memory of the Creative Partner.
Successful candidates here will have already built with agents and LLMs. They’ve spent large amounts of time working with the tools and expanding / overcoming their limitations. They are exceptionally motivated and passionate to build amazing tools, and are just looking for an opportunity to work on it in a live use case, with a blue-sky mandate and an opportunity to fulfill creativity.
Responsibilities
- Own the development of the AI Agent and role of LLMs within Luma’s main product, taking a highly creative, blue sky approach to the possibilities AI Agents open up for functionality between our users and features
- Work with design & product engineering teams to ensure the agent matches the tone and functionality that Luma is looking for
- Work with the product and infrastructure engineering team to develop architecture that scales to millions of users, and ensure latency is kept extremely low
- Work with the applied research team to understand and build for new GenAI features becoming available, including controllability & editability functions alongside new foundation models and how they can work in an agentic workflow
Background
- Experience building incredibly cool stuff with LLMs and Agents, testing and expanding their functionality beyond the industry norms. Best candidates may have built a production-level project or created an MVP to test out a startup idea
- 5+ years of engineering experience in various tech environments, including product-driven companies
- Strong python skills and deep understanding of end to end backend systems and databases
- Has prior experience working in startup environments, either as a founder or an early-stage engineer.