Platform Engineer - Memory

About Layer10

Layer10 is building the agent deployment platform — with memory as a core primitive. We believe that agents without persistent, structured memory are fundamentally limited. Our platform gives teams the infrastructure to deploy agents that remember, learn, and evolve over time. We’re a small, early-stage team working at the frontier of what agents can do when they’re built on a real foundation.

The Role

As a Platform Engineer focused on Memory, you’ll be at the center of what makes Layer10 different. You’ll work on the systems that give agents durable, queryable, and contextually rich memory — spanning vector stores, relational models, and novel retrieval architectures. This isn’t a research role; you’ll be shipping production infrastructure that real agent deployments depend on every day.

What You’ll Do

•      Design and build memory subsystems that power agent recall, context assembly, and long-term learning across deployments

•      Work hands-on with vector databases (Pinecone, Weaviate, Qdrant, or similar) to optimize storage, indexing, and retrieval at scale

•      Build and iterate on retrieval pipelines — from embedding generation to re-ranking — that serve high-quality context to agents in real time

•      Develop relational memory models that capture structured agent state, session history, and cross-agent knowledge sharing

•      Integrate with agent harnesses and SDKs (Claude Agent SDK, Codex, LangGraph, and others) to ensure memory is a first-class citizen in the agent runtime

•      Instrument memory systems with observability to understand recall quality, latency, and cost trade-offs

•      Collaborate directly with the founding team to shape the memory architecture and product roadmap

What We’re Looking For

•      3–5 years of software engineering experience, with meaningful time spent on data-intensive or ML-adjacent systems

•      Hands-on experience with vector databases and/or embedding pipelines in production (not just prototypes)

•      Strong fundamentals in Python and/or TypeScript, and comfort working across the stack when needed

•      Familiarity with agent frameworks and LLM tooling — you’ve built with at least one agent SDK or orchestration layer

•      An intuition for information retrieval: you think about precision, recall, latency, and relevance as engineering problems

•      Comfort operating in an early-stage environment where scope is broad, context shifts fast, and ownership is real

Bonus Points

•      Experience with graph databases (Neo4j, etc.) or knowledge graph construction

•      Background in search infrastructure (Elasticsearch, Solr) or recommendation systems

•      Contributions to open-source agent or memory tooling

•      You’ve thought deeply about how agents should “remember” — and have opinions about what the current approaches get wrong

Why Layer10

•      Ground floor of a company solving one of the hardest unsolved problems in agent infrastructure

•      Small team, massive surface area — your work will directly shape the product and the platform

•      Fully remote, async-friendly, built around deep work

•      You’ll work alongside a founding team that ships fast and cares about craft

Salary Range

USD 220K - 250K in base salary commensurate with experience + equity compensation