we’re on a mission to build robots that become a part of our social fabric. shapes are general purpose social agents — can be designed for anything, carry all the wide range of social relationships people have, and truly agentic. shapes reach people on the platforms they already spend time on.

You will be responsible for building and maintaining the AI infrastructure that powers each shape’s agentic capabilities.

• Proven track record of execution: You have 5+ years of engineering experience, including some time training and deploying world-class AI models
• You have hands-on production experience working with proprietary LLMs and fine-tuning open-source LLMs with HuggingFace, PEFT, RLHF, etc.
• Experience with evaluations and running hypothesis-driven experiments to benchmark the performance of LLM and RAG systems
• You have a track record of designing and implementing successful AI projects
• You will own the creation of an intelligent model routing layer to better serve messages and manage costs.
• High level of agency and self motivation, and a proven track record of developing complex systems quickly and independently.
• Enthusiasm for the dynamic pace and impact of startup environments, with a strong commitment to achieving technical excellence and innovation.

Ideal backgrounds are:

• Substantial experience with our stack: GCP, Python, Turbopuffer, Next.js, MongoDB
• Launched AI apps on your own in the past.
• Love AI and want to build the future of general purpose social agents
• Experience with discord.py is a plus
• Experience in deployment and orchestration is a plus