Senior AI Platform Engineer

About the opportunity

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. 20,400 zł - 25,900 zł a month We may use artificial intelligence (AI) tools to support parts of the hiring process, such as reviewing applications, analyzing resumes, or assessing responses and identifying potential inconsistencies or verification signals in application materials based on available information. These tools assist our recruitment team but do not replace human judgment. Final hiring decisions are ultimately made by humans. If you would like more information about how your data is processed, please contact us.

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.

Qualifications

• 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).