5 months ago
Responsibilities
- Design, implement, and deploy production AI agents using multi-step reasoning, tool calling, multi-agent coordination, and human-in-the-loop handoffs.
- Build agent harnesses covering context management, tool definitions, memory, feedback loops, observability, and lifecycle control.
- Engineer context pipelines using dynamic retrieval, re-ranking, semantic search, and GraphRAG within agentic reasoning loops.
- Implement reliability controls including retries with backoff, cost controls, structured output validation, sandboxed tool execution, and checkpoint-resume workflows.
- Develop evaluation frameworks with evals, golden datasets, regression suites, and observability traces.
- Architect scalable backend services, REST and GraphQL APIs, external integrations, database connections, and robot data-stream integrations.
- Own deployment, monitoring, observability, Docker, Kubernetes, CI/CD pipelines, and LLM-specific tracing and cost tracking.
- Build React and Next.js interfaces for robot fleet operations, engineering tools, and customer-facing applications.
- Own features from requirements and implementation through testing, rollout, and maintenance, while treating prompts as versioned production artifacts.
Requirements
- 5+ years of professional software engineering experience with a full-stack production track record.
- Strong production-level command of Python and/or TypeScript, including maintainable, testable, and performant code.
- Backend experience with Go, Rust, or TypeScript/Node.js, including RESTful and GraphQL APIs, PostgreSQL, asynchronous programming, caching, integrations, and webhooks.
- Frontend proficiency with React, Next.js, and TypeScript for production-grade user interfaces.
- Experience with automated testing, CI/CD pipelines, code review, logging, metrics, alerting, Docker, and Kubernetes.
- Hands-on experience building and deploying production LLM-powered systems or AI agents, including understanding of non-determinism, prompt injection, tool-calling loops, and cost spirals.
- Experience with at least one LLM API such as Anthropic Claude or OpenAI and agentic frameworks such as LangChain, LangGraph, or PydanticAI.
- Ability to design agent guardrails including structured output validation, retries, fallbacks, and human-in-the-loop patterns.
- Preferred experience with harness engineering, agentic retrieval, GraphRAG, hybrid search, semantic layers, durable execution, vector databases, robotics or IoT, multi-tenant platforms, AI security, and AI coding agents.
Tech Stack
Categories
About Dexmate
Dexmate builds autonomous robot platforms paired with a software stack that lets developers and enterprises build, deploy, and scale robotics applications. The company monetizes through sales of robotic hardware and licensing/support for its software platform. Founded in 2024 and headquartered in Santa Clara, California, Dexmate is privately held.