The platform uses multi-agent workflows, MCP tools, context engineering, automated guardrails, browser verification, and LLM observability to plan, develop, test, and validate software changes.

As an AI Engineer, you will help make the coding agent more accurate, reliable, and production-ready, continuously improving it through engineering, evaluation, and real-world feedback.

Key Responsibilities :

- Design and improve AI agent loops, tools, prompts, and workflows.

- Build MCP integrations for browser, development, and database tools.

- Develop guardrails, validators, linters, and safe execution policies.

- Improve context management, compaction, checkpoints, and agent reliability.

- Optimize Anthropic/OpenAI model routing for quality, cost, and latency.

- Debug agent failures using Langfuse traces and production data.

- Build golden datasets, automated tests, and evaluation pipelines.

- Ship backend, frontend, and AI platform features end to end.

Must Have :

- 2+ years in LLM/AI agents.

- Strong TypeScript / Node.js and experience with Angular, React, or Vue.

- Production experience with LLM-powered applications.

- Hands-on experience with OpenAI/Anthropic APIs, tool calling, structured outputs, and streaming.

- Experience with agent debugging, prompt/context engineering, testing, and evaluation.

- Strong SQL, Git, API, and software engineering fundamentals.

Good to Have :

- MCP, Langfuse, OpenTelemetry, LangGraph, or Playwright.

- Claude Code, Cursor, or similar AI coding tools.

- Docker, CI/CD, Linux, Kubernetes, or cloud experience.

- Laravel/PHP or AngularJS experience.

- Open-source AI projects or technical portfolio.

Tech Stack :

Node.js, TypeScript, Angular, Express, Vercel AI SDK, Anthropic, OpenAI, MCP, Playwright, PostgreSQL, Langfuse, OpenTelemetry, Docker.

The role, in one sentence :

- Make the coding agent reliably produce correct code, and build the measurement that proves it.