Memory Architect for Self-Improving AI

Letta; San Francisco (CA); On-site; USD 120,000 - 150,000; Full time

Job summary

A cutting-edge AI startup in San Francisco is seeking a professional to develop long-term memory architectures for LLMs. In this role, you will collaborate with top talent to create and improve memory systems that enhance AI capabilities. The ideal candidate will have deep expertise in LLMs, a proven track record in impactful research, and a drive to work in a dynamic environment. This opportunity demands a commitment to a full in-person presence and is geared toward those who thrive in high-stakes scenarios.

Qualifications

• Deep expertise in large language models (LLMs) and memory architectures.
• Demonstrated history of significant research contributions or open-source projects.
• Ability to work effectively on a fast-paced, dynamic team in AI.

Responsibilities

• Design the long-term memory architecture for LLMs.
• Build memory architectures supporting diverse memory types.
• Research memory sharing solutions for multi-agent collaboration.
• Enhance context management techniques to solve long-context issues.
• Run evaluations for measuring memory improvements for agents.

Skills

Deep expertise in LLMs and retrieval

Ability to balance execution speed with empirical rigor

Track record of impactful research

Job description

A cutting-edge AI startup in San Francisco is seeking a professional to develop long-term memory architectures for LLMs. In this role, you will collaborate with top talent to create and improve memory systems that enhance AI capabilities. The ideal candidate will have deep expertise in LLMs, a proven track record in impactful research, and a drive to work in a dynamic environment. This opportunity demands a commitment to a full in-person presence and is geared toward those who thrive in high-stakes scenarios.