Staff Software Engineer
Pubblicato il 03-10-2026 - Pfizer in Italia
ROLE DESCRIPTIONThe AI Backend Python Engineer (Staff) will provide technical leadership and define the architecture for backend systems that power the next generation of generative AI applications and large language model(LLM) integrations. This role focuses on driving strategic technical direction, designing platform-levelsolutions using Python, FastAPI, and AWS, and influencing engineering practices across the organization.KEY RESPONSIBILITIESDefine technical strategy and architecture for Python-based AI backend platforms (on FastAPI) that support LLM integration, generative AI pipelines, and AI agent orchestrationArchitect scalable, reliable, and secure systems on AWS infrastructure (Lambda, EKS) that serve as the foundation for multiple teamsDrive technical decisions that impact organization-wide engineering practices and platform capabilitiesDesign and champion AI system observability using LangChain, Langfuse, and litellmEstablish engineering standards for monitoring, logging, observability,
and operational excellence across AI backend servicesLead cross-functional initiatives to optimize performance, security, and scalability of large-scale AI applicationsMentor and technically guide engineers at all levels, elevating team capabilities through code reviews, architecture reviews, and technical leadershipIdentify technical debt, performance bottlenecks, and architectural limitations; drive long-term solutionsCollaborate with engineering leadership, product, and other stakeholders to align technical strategy with business objectivesInfluence technology choices, design patterns, and best practices across the AI Software & Operations departmentRepresent the team in organization-wide technical discussions and drive adoption of innovative solutionsBASIC QUALIFICATIONS7+ years of professional software engineering experience with extensive backend and distributed systems expertiseDeep expert-level proficiency in Python for building large-scale production systemsCompr