Research Engineer - Agent Memory
Core
Building the memory layer for AI agents, enabling long-term memory, context retention, and learning from interactions to distinguish chatbots from persistent agents.
Role type
Senior IC research engineer (AI memory systems)
Builds
Production-ready memory features including extraction, updates, consolidation, forgetting, and dreaming capabilities for AI agents.
Domain
Artificial Intelligence / Large Language Models / Retrieval Augmented Generation
Deliverable
production ML models
Required skills
RAG and information retrieval, LLM fine-tuning, experimental design, Python, PyTorch, vLLM, evaluation framework design, data pipeline orchestration
Preferred skills
Publications at NeurIPS/ICML/ACL, privacy-preserving ML, memory/retrieval literature expertise, vector-DB internals, deduplication, contradiction detection
Technologies
PyTorch, vLLM, Python
Responsibilities
Train models for memory extraction, updates, consolidation, and conflict resolution; prototype and benchmark paper ideas for production; build automated evaluation metrics and online A/B tests; validate hypotheses through field trials; design APIs and data contracts for safe rollouts.
Seniority
Senior, hands-on IC
