> Markdown version of [/jobs/ext/4603701-staff-software-engineer-search-retrieval-platform](https://www.wearedevelopers.com/jobs/ext/4603701-staff-software-engineer-search-retrieval-platform). Every page supports `.md` or `Accept: text/markdown`. Links point to the HTML versions so they work for humans too. Agent guide: [/agents.md](https://www.wearedevelopers.com/agents.md). --- # Staff Software Engineer, Search Retrieval Platform - **Company:** Thomson Reuters Holdings Inc. - **Location:** New York, NY, United States - **Experience:** Expert - **Salary:** $136,000.0 - $253,000.0 - **Contract:** Permanent contract - **Skills:** Java (Programming Language), Application Programming Interfaces (APIs), Encodings, Software Design Patterns, Distributed Systems, Elasticsearch, Information Retrieval, Python (Programming Language), Machine Learning, Software Engineering, Management of Software Versions, Large Language Models, Agentic-AI, Information Technology, Search Engines, Virtual Agents, Data Pipelines, OpenSearch - **Published:** November 6, 2026 - **Apply:** https://thomsonreuters.wd5.myworkdayjobs.com/External_Career_Site/job/United-States-of-America-New-York-New-York/Staff-Software-Engineer--Search-Retrieval-Platform-_JREQ205148 ## About the Role * 8+ years of software engineering experience with progression to staff-level scope, including ownership of a functional area and cross-functional leadership. A degree in Computer Science, Engineering, or a related field, or equivalent experience * Experience building search, retrieval, or data-intensive platforms at scale, with strong distributed systems fundamentals, in a language such as Python, Java, or Go * Track record translating customer needs into platform capabilities, working across product, engineering, and science * Experience designing APIs and services consumed by multiple teams, with attention to contracts, versioning, and developer experience * Operational rigor: SLOs, observability, on-call ownership, and learning from incidents * Fluency with AI-assisted development tools Preferred qualifications: * Production experience with search engines such as Vespa, OpenSearch, or Elasticsearch * Experience with hybrid (dense and sparse) retrieval, ranking, and embedding pipelines * Experience building tools or APIs for AI agents, including agentic retrieval patterns and protocols such as MCP * Search evaluation experience: relevance metrics, LLM-as-judge, and online experimentation * Multi-tenant authentication, authorization, and entitlements * Experience operationalizing ML models in production, including serving, promotion, and monitoring ## Description Advanced Content Engineering (ACE) is seeking a Staff Software Engineer to democratize retrieval at Thomson Reuters. You will build the platform capabilities that let any product team, research team, or AI agent get high-quality, customizable search without building it from scratch. That takes customer obsession: understanding what teams need, translating it into reusable platform capabilities, and making retrieval a fully customizable, plug-and-play service. You will work closely with product, research engineering, and applied science partners. The scope spans APIs, search services, agents, evaluation and grading, and authentication and security, powering end customers, AI agents, and internal systems at enterprise scale. Staff Engineers here define, build, deploy, and operate what they ship. We ship to production constantly and use AI-assisted development as the norm. Key Responsibilities: * Turn customer and product needs into platform capabilities, working directly with product teams, client teams, and agent builders to learn what they need and where the platform gets in their way. * Design the retrieval APIs and search services that serve end users, AI agents, and internal systems, with consistent contracts, tunable filtering and ranking, predictable latency, and result payloads built for agent context windows. * Make retrieval customizable by design: pluggable components for chunking, enrichment, embedding, ranking, and custom models that teams can compose and tune through self-service APIs. * Own how content is ingested and indexed on the platform (streaming pipelines, search engine configuration, zero-downtime index changes), so customization reaches all the way down the stack. Partner across disciplines * Partner with Research Engineering and Applied Science to take new retrieval techniques, such as embedding models, hybrid retrieval, re-ranking, and agentic patterns, from experiment to supported platform capability, with a clear path to production. * Work with product and client teams to prioritize the roadmap, bring adoption feedback back into the platform, and reduce time-to-value for new adopters. Quality, security, and trust * Build evaluation into the platform: offline evaluation, search grading, LLM-as-judge scoring, and online experiments, so every change to retrieval can be measured. * Own the security posture of retrieval: multi-tenant data isolation, entitlements enforced at query time, and telemetry that protects user content. * Own SLOs, observability, and on-call for what you build, including tracing for agentic retrieval paths, and drive improvements from incident reviews. Technical leadership * Set architectural direction and design patterns across retrieval, search services, and the APIs teams build on, and act as the technical authority others turn to. * Lead cross-team initiatives, set priorities, and make hard trade-offs between near-term product needs and the long-term platform. * Mentor Senior and mid-level engineers, and help the team use AI-assisted development effectively.