Agentic Software Engineer (Full Stack)

Specright

What You'll Do

• Build and Ship Agentic Applications Design, develop, and deliver agentic products end-to-end - including chat-based interfaces, AI workflow tools, and internal copilots - that serve real users in production.
• Leverage Agentic Coding Workflows Use modern agentic coding tools daily (e.g., Claude Code, Codex, Cursor, and similar assistants) to accelerate delivery while maintaining high code quality and engineering standards.
• Design Agent Systems Architect agent internals: tool calling, multi-step orchestration, memory and state management, retrieval pipelines, permissions models, and failure recovery strategies.
• Own Evals and Quality Build and maintain agent evaluation infrastructure - regression suites, scenario-based tests, golden traces, and both offline and online metrics - so you always know whether your agent got better or worse.
• Implement Observability Instrument agents with production-grade observability: distributed tracing, prompt and version tracking, tool-call telemetry, and cost/latency monitoring.
• Own Production Readiness Take responsibility for the full lifecycle: rollout plans, QA, guardrails, incident response, and continuous improvement of deployed agents.

What You Bring (Must-Haves)

• Proven full-stack engineering ability. You've shipped production software end-to-end - not just prototypes or demos.
• Hands-on experience building LLM-powered agents that work in production. You've owned at least 1 2 agentic features or products through deployment and operation.
• Deep familiarity with agentic coding best practices. You understand task decomposition, agent delegation, prompt hygiene, iterative loops, and how to review and steer agentic outputs effectively.
• Practical experience with agent evals and tracing. You've gone beyond model benchmarking to build evaluation systems specifically for agent behavior - including trace-based debugging and regression detection.
• Solid engineering fundamentals. Testing, debugging, security basics, performance optimization, and reliability are second nature to you.

Nice-to-Haves

• Proficiency in TypeScript/Node.js and/or Python, with experience in modern web stacks and startup-pace delivery.
• Familiarity with agent tooling ecosystems: prompt/version stores, eval harnesses, tracing and telemetry platforms (OpenTelemetry, LangSmith, Datadog, Honeycomb, or similar).
• Experience with RAG architectures: vector databases, indexing strategies, chunking approaches, and retrieval evaluation.
• Experience designing tool APIs for agents: structured output schemas, idempotency, rate limiting, and error handling patterns that make agents more reliable.