Senior AI Engineer (Agentic AI)
Posted Updated
Description
- Thomson Reuters is seeking a Senior AI Engineer, Agentic AI to build and ship LLM-powered agents for our go-to-market organization. Our agent platform brings the tools sales teams use today into one AI-agent hub. You will build specialist agents that are scalable, reliable, and secure, and help set the engineering patterns the team relies on as it grows
- Design, develop, and deploy production LLM agents end to end, including multi-step workflows, retrieval-augmented generation (RAG), prompt design, tool and function calling, and structured-output validation
- Build specialist agents that ingest data from enterprise systems, extract actionable signals, and recommend next steps through a human-in-the-loop review
- Build and operate evaluation and observability that make agents trustworthy: guardrails, automated evals, drift detection, and production monitoring
- Integrate agents with AWS infrastructure and enterprise APIs, including MCP servers where approved
- Contribute to shared agent patterns so new agents reuse common context, guardrail, and evaluation components rather than rebuilding them
- Craft and sustain design documentation, architecture decision records, and system models
- Collaborate with platform, data integration, and governance engineers, plus product stakeholders, to turn business requirements into robust designs
- Provide technical guidance and code and design reviews, and mentor engineers as the team builds its agentic capability
- SQL and data integration experience, including working with cloud data warehouses and enterprise APIs
- At least 2 years of experience delivering practical applied AI or machine learning solutions that improve user efficiency and reliability
- A working understanding of the machine learning lifecycle (feature engineering, training, evaluation, monitoring), sufficient to collaborate with data scientists and ML engineers
- Experience with AWS services (such as ECS, Bedrock, SageMaker, or Secrets Manager), including networking and security principles
- Practical experience building agent workflows with guardrails, evaluation, and reliability in mind
- A creative, resourceful, and effective problem-solving approach
- Experience deploying services with Docker, CI/CD pipelines, and infrastructure as code
- Hands-on experience building production LLM applications, including RAG, prompt engineering, and tool or function calling
- Strong Python programming skills, with modular, well-tested, production-grade code
- Experience leading or supporting engineers through code and design reviews
- Clear technical communication, with a bias toward information sharing and diverse viewpoints
- Effective use of AI coding assistants (e.g., Claude Code, Cursor, Copilot) in a team setting
- Experience with the Python ML stack (e.g., scikit-learn, PyTorch, Pandas, XGBoost/LightGBM) and data validation frameworks (e.g., Pandera, Great Expectations, or Pydantic)
- Familiarity with responsible AI practices, including hallucination mitigation, privacy, governance, and audit logging
- Knowledge of vector stores, embeddings, and retrieval pipelines
- Experience with MCP servers and gateways for enterprise tool integration
- Experience integrating with enterprise sales or GTM platforms (such as Salesforce, Gong, or Conga), or familiarity with sales concepts like opportunity management, pipeline analytics, or MEDDPICC
- Experience designing multi-agent systems, including orchestrator and sub-agent patterns, planning, tool use, and memory
- Hands-on experience with agentic frameworks such as LangGraph, Strands, LangChain, or CrewAI
- End Date: June 30, 2027
Skills
- Agentic AI
- AI
- Analytics
- API
- AWS
- CI/CD
- Claude Code
- Cloud
- CrewAI
- Docker
- ECS
- Embeddings
- Feature Engineering
- Go-to-Market
- Infrastructure as Code
- LangChain
- LangGraph
- LLM
- Machine Learning
- MCP
- Networking
- Observability
- pandas
- Prompt Engineering
- Pydantic
- Python
- PyTorch
- RAG
- SageMaker
- Salesforce
- scikit-learn
- SQL
- XGBoost
Sign in to read and join the discussion.