platform engineer in London

Описание

Пleo builds spend management solutions that help finance teams and employees manage business spending more seamlessly and effectively.

Задачи

• Design, build, and operate core Gen AI platform components, including an LLM routing gateway, vector search and RAG infrastructure, tool registry and MCP gateway, AI observability and evaluation tooling, and infrastructure for long-running agentic workflows;
• Own production-quality delivery of platform features from design through rollout, monitoring, and follow-up;
• Contribute to resilient system design with sensible APIs, failure handling, rate limiting, retries, idempotency, and safe change management;
• Improve reliability and observability through metrics, dashboards, alerting, incident follow-ups, and operational improvements;
• Partner with Applied AI Engineers and product teams to understand platform needs and help them build AI-powered features safely;
• Build internal SDKs, templates, and guardrails for product engineers;
• Support other engineers through pairing, code reviews, technical feedback, and clear documentation;
• Help evaluate build-versus-buy decisions in the LLMOps tooling landscape;
• Develop a clear picture of how AI features are built at Pleo and identify infrastructure bottlenecks;
• Take ownership of a core platform component and improve its reliability, observability, or developer experience;
• Deliver production-ready improvements with rollout plans, monitoring, and operational documentation;
• Partner with Applied AI Engineers and product teams to identify platform investments;
• Contribute to Pleo's internal standards for AI feature development, quality evaluation, prompt management, and production monitoring.

Требования

• Strong backend and systems engineering background with experience building and operating production services with reliability and observability requirements;
• Experience designing and delivering shared platform or infrastructure components used by multiple teams;
• Strong production ownership, including monitoring, alerting, incident response, debugging, and post-incident learning;
• Knowledge of distributed systems fundamentals, including asynchronous workflows, idempotency, consistency trade-offs, and designing for failure;
• Hands-on experience with LLM APIs or strong interest in learning about rate limits, context windows, multi-vendor routing, latency variance, and cost control;
• Security mindset for AI systems, including prompt injection risks, PII in logs, data leakage, and safe credential handling;
• Strong programming experience in a JVM-based language or Python, with the ability to contribute to Kotlin and Python components;
• Clear communication and collaboration skills for turning ambiguous platform needs into practical solutions;
• Passion for developer experience and enabling engineering teams;
• Deep understanding of LLMOps, data retrieval, prompt and context engineering, and model evaluation in production;
• Ability to work across languages and technologies to achieve goals;
• Ability to explain AI trade-offs clearly to non-technical stakeholders;
• Proven experience building platforms or tooling for agentic AI;
• Ability to work fully autonomously on an entire product feature from design to implementation;
• Candidates primarily interested in model research or algorithm development are not a fit;
• Candidates who prefer building customer-facing features are not a fit.

Условия

• Remote, hybrid, or in-person work options are available depending on the location;
• The employee must be physically based in the chosen country with a valid right to work;
• Visa sponsorship is not available;
• Pleo card provided;
• Catered meals or a lunch allowance is available for work days;
• Comprehensive private healthcare is provided depending on location;
• 25 Days of holiday plus public holidays;
• Free mental health and well-being support through Mynd Up;
• Paid parental leave;
• The interview process includes an intro call, technical screening, system design interview, live coding interview, Hiring Manager interview, and final leadership interview;
• #J-18808-Ljbffr