Staff Engineer – Python Agent Engineer

Job Purpose Why This Role Exists  T2's intelligence lives in its agent layer — five distinct AI agents (Assistant, Content Analysis, Planning, Quality Evaluation, and Transformation) implemented in Python + LangGraph, each deployed as an independent service on EKS. These agents execute the Planning→Generation→Validation loop that is T2's core differentiator: the system that reasons about content, builds dynamic transformation plans, iterates on quality until exit conditions are met, and adapts mid-flight when quality gates fail.  The current team has strong AI/LLM understanding and architectural thinking, but their primary output is the Java orchestration backbone and the LLM provider/prompt landscape. Nobody's primary job today is delivering production-grade Python agent services — and the agent layer is T2's critical path to MVP. About Transform At RWS, we enable the world's largest enterprises to communicate with global audiences through cutting-edge language technology, AI-driven solutions, and expert services. Our RWS Transform division empowers organizations to accelerate digital transformation, scale global content, and unlock growth in every market. Job Overview What This Person Owns Agent services as production software. Not prototypes, not notebooks — containerized Python services with proper error handling, retry logic, graceful degradation, structured logging, and health checks. Each agent follows the common blueprint/library pattern (per ADR) sharing auth, observability, and messaging patterns.  The Planning→Generation→Validation execution loop. This is the heart of T2. The Planning Agent takes content analysis results, governance rules, and user requirements to produce a dynamic, mutable plan. During execution, agents can add, modify, or reorder steps based on real-time quality assessment. The engineer owns making this loop reliable, performant, and observable — including the exit conditions (human escalation, quality objectives met, error/timeout).  LangGraph orchestration graphs. Each agent has its own state machine defining how it reasons, what tools it invokes, and how it handles branching logic (e.g., Content Analysis detecting PII triggers a different downstream path than detecting marketing content). These graphs need to be well-structured, testable, and maintainable. Quality Evaluation Agent specifics. The multi-level quality assessment (segment → file → cross-file → persona validation) with the Quality × Confidence scoring model. This includes the iterative improvement loops bounded by token budgets, and the Red Team adversarial testing layer. What This Person Is Not Not a data scientist or ML researcher. Not someone who fine-tunes models.This is a software engineer who builds reliable distributed services that happen to orchestrate LLM calls.  Technical Profile  • Strong Python (3.12+, async/await, type hints, FastAPI)  • LangGraph / LangChain — practical experience building multi-step agent workflows, not just tutorials  • Production service patterns — Docker, health checks, structured logging, graceful shutdown, circuit breakers  • Testing agentic systems — deterministic testing strategies for non-deterministic outputs, mock LLM responses, integration test patterns  • Message queues — RabbitMQ consumer/producer patterns for async agent coordination  • Nice to have: AWS Bedrock experience, familiarity with translation/NLP domain, experience with multi-model routing  Collaboration Pattern  Works most closely with the 2 senior backend engineers (who own the Java orchestrator that invokes agents) and the GenAI-native developer (who owns LLM provider integrations, prompt management, and BYOLLM configuration). Takes architectural direction from the architect on agent boundaries and data contracts.  Life at RWS Life at RWS – If you like the idea of working with smart people who are passionate about growing the value of ideas, data and content by making sure organizations are understood, then you'll love life at RWS.  Our purpose is to unlock global understanding. This means our work fundamentally recognizes the value of every language and culture. So, we celebrate difference, we are inclusive and believe that diversity makes us strong. We want every employee to grow as an individual and excel in their career.  In return, we expect all our people to live by the values that unite us: to partner, putting clients fist and winning together, to pioneer, innovating fearlessly and leading with vision and courage, to progress, aiming high and growing through actions and to deliver, owning the outcome and building trust with our colleagues and clients. RWS embraces DEI and promotes equal opportunity, we are an Equal Opportunity Employer and prohibit discrimination and harassment of any kind. RWS is committed to the principle of equal employment opportunity for all employees and to providing employees with a work environment free of discrimination and harassment. All emp