Agent Platform Engineer

Replit•Foster City, CA

19h•Hybrid

About The Position

In this role you will build powerful tools that help product engineers iterate rapidly on the Agent experience and directly enhance the core Agent itself. You’ll bridge the gap between the AI team (working on the core Agent logic) and the UX team (crafting delightful Agent experiences), enabling both groups to excel within their specialties. This role blends systems engineering, developer experience and product engineering. We tackle complex challenges across the full stack, from browser-based interfaces to high-performance backends to Linux systems engineering. We’re looking for engineers who have a keen sense of the product experience and how to power it with performant systems. On this team, you’ll have the opportunity to grow your skills across our infrastructure and product, and to lead end-to-end efforts with meaningful impact. We value diverse perspectives and encourage candidates from all backgrounds and experiences to apply.

Requirements

5+ years professional experience in building backend services, systems and platforms
Someone who values simplicity and pragmatism—you think radically but ship incrementally.
Self-directed and comfortable working autonomously.
Collaborative and team-oriented with a strong sense of shared ownership.

Nice To Haves

Experience across the full-stack.
Experience building AI powered tooling or products.
Experience with collaborative editing technologies such as Operational Transforms (OT), CRDTs, or the inner workings of Git.
Experience with event sourcing systems.

Responsibilities

Build high-throughput backend applications and services, like streaming chat between user and agent.
Design a collaborative "Multiplayer Computer" that lets humans and AI agents work together on shared shells, filesystems, and state—conflict-free and in real time.
Develop infrastructure (frontend & backend) that empowers product engineers to rapidly ship delightful user experiences.
Design and build durable, well-crafted interfaces between the Agent and internal systems, as well as external agentic systems (MCP)