Drimva

ByteDance

Senior Software Engineer - AI Agent Memory Infrastructure

San Jose, California · Senior · Full-time · Posted Oct 28, 2026

Education
Bachelor’s degree or higher in Computer Science, Artificial Intelligence, Data Science, or a related field.

About the job

Builds and operates scalable, low-latency memory infrastructure that enables AI agents to retain, retrieve, and use long-term, conversational, task-oriented, and multimodal information.

What you'll do

  • Design and evolve unified AI agent memory platforms and their data pipelines, including ingestion, storage, indexing, retrieval, updating, compression, and forgetting.
  • Develop memory representations and workflows for retrieval, ranking, conflict resolution, summarization, and multimodal fusion.
  • Partner with model, application, and platform teams to productionize memory capabilities and optimize quality, latency, cost, reliability, and safety.

What they ask for

  • Bachelor’s degree or higher in Computer Science, Artificial Intelligence, Data Science, or a related field.
  • Strong experience in distributed systems, databases, information retrieval systems, or AI infrastructure, including system design and production engineering.
  • Proficiency in at least one of Go, Python, or C++, with strong coding standards and engineering practices.
  • Understanding of LLM application technologies and familiarity with at least one key area of memory systems.

Nice to have

  • Experience with agent memory systems, user profiling, recommendation/search platforms, or knowledge bases; contributions to or deep understanding of open-source memory frameworks such as mem0, memOS, or memU.
  • Experience in multimodal processing, online inference, personalized agents, or long-term user state modeling; impactful publications, open-source work, or technical achievements.

Tools

Go, Python, C++, LLMs, embeddings, retrieval-augmented generation (RAG), vector/graph indexing

AI focus

Develops AI agent memory systems for long-term personalization, conversational and task-oriented memory, multimodal fusion, and real-time inference.

Model-written summary · the original posting is authoritative

First seen Oct 28, 2026 · Last verified Nov 5, 2026 (UTC)