"JOB SUMMARY": "Our team is building an AI agent that helps users design and build applications on our platform, and we're looking for an engineer to help shape it - from how the agent reasons, plans, and recovers from failure, to the tools it uses to build and test applications, to the backend that keeps it fast and dependable at scale. This is a blend of applied AI engineering and backend systems work, well suited to someone who has built agentic products and enjoys owning ambiguous, high-impact problems end to end. You'll work across a TypeScript backend (Node.js / Express) and a Python (FastAPI) AI service that orchestrates the agent.",
"Key Responsibilities":
"Agent reasoning & behavior - Design and refine how the agent interprets intent, plans multi-step work, calls tools, and recovers from failure - across an orchestrator and specialized builder, planner, tester, and searcher agents.",
"Agent tooling - Build and harden the tools the agent uses to generate, run, and test applications on our platform, including self-healing tool retries and safe execution.",
"Backend services at scale - Develop and operate the services that power the agent in Node.js / TypeScript (Express) and Python (FastAPI), with a focus on performance, reliability, and observability.",
"LLM integration - Work with Anthropic Claude (via Vertex AI / LiteLLM) and Gemini, applying context caching, conversation compaction, and Model Context Protocol (MCP) tooling.",
"APIs & real-time streaming - Build RESTful APIs and Server-Sent Events (SSE) streams that connect the agent, the backend, and the front end.",
"Data & persistence - Model and optimize data in MongoDB (Mongoose / PyMongo), with Redis for caching and performance.",
"Quality & evaluation - Build the evaluation, instrumentation, and feedback loops that measure and improve agent quality, latency, and cost over time.",
"Testing & reliability - Write and maintain unit, integration, and end-to-end tests (Jest, Playwright, pytest) and contribute to CI/CD on Google Cloud.",

"Required Qualifications":
"7+ years of professional software engineering experience, with substantial backend work.",
"Strong software engineering fundamentals - you design clean systems, reason about tradeoffs independently, and ship well-tested code that holds up in production.",
"Experience building and operating backend services at scale, with a feel for performance, reliability, and observability; proficiency in TypeScript and Python.",
"A track record of building AI- or agent-powered products, and good judgment about how to apply LLMs to real problems - what they're good at, where they break, and how to design around both.",
"Strong product intuition - you make sound decisions in ambiguity, weigh user needs against technical tradeoffs, and know which problems are worth solving.",
An avid user of AI coding tools, energized by exploring how AI can automate workflows, enhance creativity, and increase your personal impact."

Preferred Qualifications:
Experience with agent frameworks or developer kits (e.g. Google ADK) and the broader LLM tooling ecosystem (MCP, CopilotKit).",
Background in developer tools or low-code / no-code platforms."
Front-end development skills with TypeScript and React (React 19 / Next.js)."
Familiarity with event-driven, real-time systems (Server-Sent Events) and high-throughput services."
Knowledge of web-application security best practices and OWASP principles."
Hands-on experience with MongoDB, Redis, and Google Cloud Platform (Vertex AI)."