Principal Engineer - Agentic AI Applications

Serko builds business travel and expense technology. The Principal Engineer will lead the technical vision and architecture for agentic AI applications, creating reliable, high-availability systems and partnering with Product, Research, Platform, and AI Ops teams.

Responsibilities:

• Architect the Agentic Runtime: Own the end-to-end technical architecture for agentic systems, including multi-agent orchestration, memory management, and reasoning pipelines
• Engineering for Reliability: Implement robust infrastructure to handle non-deterministic loops, including automated circuit-breakers, latency-aware orchestration (targeting P99 SLAs), and granular token/cost-observability
• Set the Standard: Define engineering patterns, design standards, and best practices for agentic application development across the entire organization
• Scale the Team: Mentor and grow Senior engineers; conduct deep-dive architecture reviews and drive high-stakes technical decision-making and be instrumental in the hiring process
• Bridge the Gap: Partner with AI Ops to ensure agentic workloads are efficiently built, served, and monitored with state-of-the-art observability

Requirements:

• 12+ years of software engineering experience, with at least 4+ years in technical leadership or principal-level roles
• Deep hands-on experience shipping LLM-powered or agentic systems to production (utilizing frameworks like LangGraph, AutoGen, CrewAI, or custom-built runtimes)
• A proven track record of architecting distributed, high-availability systems
• Mastery of Python and modern backend frameworks
• Deep familiarity with RAG pipelines, vector databases, tool-calling APIs, and multi-step reasoning
• Ability to distill complex "black box" AI concepts into clear technical roadmaps for executive stakeholders and cross-functional partners
• A desire to influence org-wide engineering culture, prioritizing code quality, observability, and rigorous testing in an AI-first world
• TypeScript/Node.js is a plus