Staff Engineer (Karlo Platform)

Join Kargo, a leading ad-tech company, as a Staff Engineer on the Karlo Platform team. In this role, you will own the shared infrastructure that powers our conversational AI agents, ensuring it is reliable, scalable, and well-documented. You will design and deliver a shared evaluation framework, ship production-ready shared services, and own checkpointing and agent context memory. You will also build SDKs, documentation, and onboarding tooling to improve the developer experience. The ideal candidate will have 7-10 years of software engineering experience, with a strong background in AI and agentic systems.

• Posséder et renforcer l'infrastructure d'hébergement des agents basée sur LangGraph, permettant à chaque équipe d'agents de construire et de déployer en toute confiance.
• Concevoir et livrer un cadre d'évaluation partagé permettant aux équipes d'agents de valider la correction, de mesurer la qualité et de détecter les régressions avant qu'elles n'atteignent les clients.
• Construire les SDK, la documentation et les outils d'intégration qui réduisent le temps d'adaptation d'une nouvelle équipe d'agents de plusieurs semaines à quelques jours.

• Experience building shared platform infrastructure consumed by multiple teams, with a track record of improving developer experience through SDKs, documentation, or internal tooling; B.S. or higher in Computer Science, Engineering, Mathematics, or equivalent practical experience; AdTech or programmatic experience a plus
• Deep hands-on experience with LangGraph in production: graph construction, node and edge design, conditional routing, state management, and agent lifecycle management; plus the broader LangChain ecosystem and LangSmith for tracing, evaluation, and dataset management
• Command of agentic design patterns — ReAct, Plan-and-Execute, multi-agent orchestration, human-in-the-loop workflows, and tool use — with experience operating multi-agent systems and state handoff across agent boundaries
• Experience with LLM evaluation methodologies and production observability for AI systems, including the ability to debug non-deterministic agent behavior methodically
• Strong LLM foundations: prompt engineering, context window management, memory architectures for multi-turn agents, token and cost optimization, RAG architectures (vector stores, embedding pipelines, retrieval, reranking), and model selection tradeoffs
• 7–10 years of software engineering experience, including at least 2 years building production AI or agentic systems — shipped and operated, not prototyped
• Strong backend fundamentals across distributed systems, API design, and event-driven architecture; TypeScript required and Golang a strong plus; cloud infrastructure (AWS, GCP, or Azure), containerization, and CI/CD