高级 AI 平台工程师
Senior AI Platform Engineer
XTB是一家来自金融行业的全球公司，专注于金融工具的在线交易。我们是波兰最大的FinTech公司，在中欧和东欧地区处于领先地位，我们的业务范围覆盖多个国家，包括亚洲和南美洲。在XTB，我们注重员工的发展，为他们提供在各个领域获取知识和技能的机会，并提供多种培训和发展计划。如果您正在寻找挑战，并希望在国际商业环境中获得宝贵的经验，XTB就是您的理想之选。
我们是一家经过认证的Great Place to Work公司。
我们正在寻找一位具备强大平台工程和DevOps专业知识的高级AI平台工程师，负责设计、构建和维护一个前沿的开发者平台，使团队能够大规模构建和部署AI驱动的解决方案。该职位将软件工程卓越性与AI工具编排相结合，重点是创建用于LLM集成、代理工作流和企业AI基础设施的自助服务平台。
职责
- 设计并实现基于LLM的路由、智能模型选择、回退策略和成本优化的企业AI平台
- 构建和维护MCP（Model Context Protocol）网关基础设施，实现企业工具和内部数据与AI代理的无缝集成
- 构建基于LangGraph的编排框架，用于复杂代理工作流，使开发者能够组合多步骤AI应用
- 实现RAG（检索增强生成）平台组件：向量搜索集成、提示优化和分块策略
- 使用Langfuse和其他类似工具建立全面的可观测性：追踪、成本监控、令牌使用分析和LLM质量指标
- 设计和优化LLM交互API，使开发者能够无缝访问平台功能
- 使用Kubernetes（K8s）和ArgoCD构建和编排流水线，实现基于GitOps的AI服务部署
- 使用GitLab CI/CD建立CI/CD基础设施，实现AI平台组件的自动化测试、部署和版本控制
- 为AI端点实现安全、身份验证和限流，防止滥用并确保多租户隔离
- 创建自助开发者界面和文档，使非AI专家也能构建复杂的代理应用
- 设计
查看英文原文
XTB is a global company from the financial industry, focusing on online trading of financial instruments. We are the largest FinTech in Poland and a leader in Central and Eastern Europe, and the range of our operations covers several countries, including Asia and South America. At XTB, we focus on the development of our employees, giving them opportunities to gain knowledge and skills in various fields, as well as offering a number of training and development programs. If you are looking for challenges and want to gain valuable experience in an international business environment, XTB is the right place for you.
We are a certified Great Place to Work company.
We are seeking a Senior AI Platform Engineer with strong platform engineering and DevOps expertise to design, build, and maintain a cutting-edge developer platform that empowers teams to build and deploy AI-driven solutions at scale. This role bridges software engineering excellence with AI tool orchestration, focusing on creating self-service platforms for LLM integration, agentic workflows, and enterprise AI infrastructure.
Responsibilities
- Design and implement an enterprise AI platform with LLM-based routing, intelligent model selection, fallback strategies, and cost optimization
- Build and maintain MCP (Model Context Protocol) gateway infrastructure enabling seamless integration of enterprise tools and internal data with AI agents
- Architect LangGraph-based orchestration frameworks for complex agentic workflows, enabling developers to compose multi-step AI applications
- Implement RAG (Retrieval-Augmented Generation) platform components: vector search integration, prompt optimization, and chunking strategies
- Establish comprehensive observability using Langfuse and similar tools: tracing, cost monitoring, token usage analytics, and LLM quality metrics.
- Design and optimize APIs for LLM interactions, enabling developers to access platform capabilities seamlessly.
- Build deployment and orchestration pipelines using Kubernetes (K8s) with ArgoCD for GitOps-driven AI service deployments.
- Establish CI/CD infrastructure using GitLab CI/CD for automated testing, deployment, and versioning of AI platform components.
- Implement security, authentication, and rate limiting for AI endpoints protecting against abuse and ensuring multi-tenant isolation.
- Create self-service developer interfaces and documentation enabling non-AI-experts to build sophisticated agentic applications.
- Design prompt management and versioning systems, enabling teams to collaborate on prompts and track performance across versions.
- Lead technical initiatives, conduct code reviews, mentor engineers, and foster AI-first engineering culture.
Requirements
- 4+ years as a Platform Engineer, Backend Engineer, or AI Engineer with hands-on platform/service delivery experience.
- Strong software engineering fundamentals: proficiency in Python and/or Go/Java, clean code practices, system design.
- Production experience with container orchestration (Kubernetes) and ArgoCD for GitOps deployments.
- Expert-level knowledge of GitLab CI/CD: designing reliable pipelines, managing deployments, optimizing build efficiency.
- Practical experience with LLM frameworks and agentic systems (LangGraph, Pydantic AI, or equivalent).
- Working knowledge of RAG systems: vector databases, similarity search, prompt engineering for retrieval-augmented contexts.
- Understanding of LLM routing and optimization: cost-aware model selection, latency reduction, multi-model orchestration.
- Experience building or managing gateway/proxy infrastructure (API gateways, reverse proxies, load balancing).
- Strong foundation in infrastructure-as-code, version control (Git), and CI/CD best practices
- Excellent communication: ability to work with AI/ML teams, data engineers, and product managers to translate requirements into platform solutions
Nice to have
- Experience building prompt management and versioning systems.
- Knowledge of semantic caching or prompt optimization techniques.
- Experience with multi-tenant SaaS architecture or platform-as-a-service (PaaS) design.
- Familiarity with evaluation frameworks for LLM outputs or agentic behavior.
- Experience with vector database optimization or semantic search.
- Knowledge of advanced deployment patterns (canary, blue-green, A/B testing) for AI services.
What we offer
- Real influence on the development of the company and the product.
- Work in an experienced team that is happy to share its knowledge.
- A clear vision of development thanks to regular feedback and clear career paths.
- Regular team-building meetings.
Benefits
- A training budget for courses and conferences that interest you.
- An extra day off on your birthday.
- An extra day off for parents.
- Equipment tailored to your needs.
- Private medical care and group insurance.
- Access to an e-learning platform for learning English and a benefits platform.
- Access to a wellbeing platform and the opportunity to take advantage of workshops and private therapy sessions.
- Remote work, from the office in Warsaw or from a coworking space in your city.
Originally posted on Himalayas