Staff Software Engineer
Thomson Reuters1 day ago
Base Salary
$118k - $220k/yr
Responsibilities
- Lead technical strategy, architecture, experimentation, testing, rollout practices, and operational excellence for AI-agentic legal software.
- Design and build backend services with Python, FastAPI, and PostgreSQL for real-time AI-agentic interactions, workflows, LLM orchestration, and content pipelines.
- Define service boundaries, system interfaces, and scalable design patterns for the AI platform.
- Support front-end engineers delivering an agentic UI using SSE, AG-UI, Claude Agent SDK, and AgentCore.
- Evaluate and adopt AI techniques, oversee proof-of-concepts, assess technical feasibility, and translate AI research into roadmaps.
- Mentor software engineers and influence architecture decisions, technical specifications, and senior stakeholders.
- Integrate OpenAI and Anthropic capabilities with orchestration, routing, error handling, guardrails, telemetry, and observability.
- Improve testing, monitoring, security, compliance, CI/CD pipelines, and DevOps automation for cloud-native production systems.
- Partner with Labs, product managers, designers, and other engineering leaders to turn early AI concepts into production-ready features.
Requirements
- 7+ years of software engineering experience, including 2+ years in technical leadership and 3+ years building scalable cloud applications, APIs, and AI-enabled products.
- Proven experience shipping production-grade AI systems, including deployment, monitoring, evaluation, scaling, and ongoing operational ownership in enterprise environments.
- Strong proficiency in Python, FastAPI, PostgreSQL, and AWS or Azure cloud services, with deep knowledge of distributed systems, architecture, and data modeling.
- Experience building and owning CI/CD pipelines for production web applications, including automated testing and safe release practices.
- Strong leadership, mentoring, strategic thinking, problem-solving, communication, presentation, product sense, and ability to work with ambiguous requirements.
- BA or BS in Computer Science or equivalent practical experience.
- Knowledge of vector databases, embeddings, semantic search, and information retrieval.
- Hands-on experience with AG-UI or comparable agent-UI protocols and SDKs, and strong proficiency with SSE events.
- Backend streaming experience with Node or Python/FastAPI sufficient to debug end to end.
- Experience with Datadog, CloudWatch Log Insights, and Braintrust.
- Experience working on legal, compliance, or other regulated professional-domain products is preferred.
Benefits
- Flexible hybrid work model for office-based roles, with work-from-anywhere flexibility for up to 8 weeks per year.
- Flexible work-life policies, career development and skills programming, tuition reimbursement, and employee incentive programs.
- Comprehensive benefits including health, dental, vision, disability, life insurance, retirement savings with company match, vacation, sick and safe paid time off, holidays, parental leave, and sabbatical leave.
- Two company-wide mental health days, access to Headspace and an Employee Assistance Program, fitness reimbursement, and wellbeing resources.
- Additional programs include commuter benefits, adoption and surrogacy assistance, flexible spending and health savings accounts, a 529 plan, legal identity theft protection, and an employee stock purchase plan.
- Two paid volunteer days annually and opportunities for pro-bono consulting and ESG initiatives.
Tech Stack
Categories
About Thomson Reuters
Thomson Reuters builds research platforms, workflow software, and data services for legal, tax and accounting, compliance, and government professionals, and operates the Reuters global news service. Flagship products include Westlaw for legal research and ONESOURCE and Checkpoint for tax and accounting, sold primarily via subscriptions and enterprise licenses. A public company headquartered in Toronto, it was formed in 2008 by combining Thomson Corporation and Reuters Group and is listed on the TSX and Nasdaq as TRI.