Senior Principal Software Development Engineer

Travelport provides technology used to book flights, cars, and hotels. The Senior Principal Software Development Engineer will lead the strategy, development, and adoption of engineering platforms, tools, and delivery capabilities while driving AI-native software delivery, engineering excellence, platform standardization, and technical mentorship.

Responsibilities:

• Work directly in the code — designing, building, refactoring, reviewing — to set the technical standard by example, not by decree
• Lead a platform/enablement squad that creates reusable capabilities and paved paths other teams adopt to move faster with less risk
• Spread patterns by shipping reference implementations teams can copy, not policies they must comply with, partnering with the squad's Director who owns day-to-day delivery
• Design, build, and operate MCP (Model Context Protocol) servers and the agent tooling and services that hang off them, exposing model-backed capabilities to product surfaces and partner/agent consumers
• Design clean APIs across both AI-native / LLM-facing and traditional REST patterns, and own the production serving path plus the feedback-loop plumbing that routes production signals back to Data Science
• Build observability and evaluation for AI systems — tracing, monitoring, and eval harnesses — as first-class platform capabilities, and advance continuous delivery by separating deployment from release (progressive rollouts, feature flags, telemetry-driven validation)
• Drive simplification before migration — consolidating services whose boundaries exist only because organizational boundaries once did — and generate high-value tests from real production behavior
• Raise the whole team's throughput with AI in the development loop — agentic coding, AI-assisted testing and code review — and establish the patterns for using these tools effectively and responsibly
• Grow strong engineers into stronger ones through pairing, code review, and everyday coaching — raising the team's confidence with MCP, AI-native API design, and production AI tooling
• Operate with "no private knowledge": everything you build is documented, observable, and survivable beyond any single person
• Make cost visible and build cost literacy into the team's decisions; define and improve the metrics that matter — delivery speed, quality, reliability, developer productivity — through engineering, not reporting

Requirements:

• 10+ years designing, building, deploying, and operating production software, with deep, current, hands-on Java plus Python and/or TypeScript for AI/MCP work (polyglot)
• Demonstrated hands-on experience building MCP servers and agent / tool-use orchestration, with strong AI-native / LLM-facing API design alongside REST
• Experience integrating and serving model-backed capabilities in production — inference APIs, serving paths, and the feedback loops that route production signals back to Data Science — with production-grade observability and evaluation for AI systems
• Enough ML literacy (model behavior, evaluation, inference trade-offs) to partner deeply with Data Science; model training itself is not required
• Mastery of modern cloud-native engineering: Kubernetes/containers, CI/CD, and production observability
• A track record of building things that made other engineers materially more productive, and of leading through technical authority and mentorship as an IC — not by managing an organization
• Depth with agent frameworks, tool-use orchestration, and LLM application patterns
• Data streaming/pipeline knowledge (e.g., Kafka) sufficient to build feedback loops; deeper data-lake tooling a plus
• Internal developer platform / developer-experience work at scale; service consolidation in large codebases; cloud cost governance / FinOps sensibility