Thomson Reuters is revolutionizing how legal professionals work through cutting-edge generative AI solutions and aims to develop next generation CoCounsel product to transform the legal industry.
Requirements
Advanced proficiency in Python, including experience developing backend APIs (preferably using FastAPI; otherwise, Flask, or Django) and integrating with external APIs and data sources.
Demonstrated experience building and consuming RESTful APIs and working with authentication, error handling, and data serialization.
A good understanding of the latest industry trends and frameworks such as the Model Context Protocol (MCP), Agent to Agent (A2A), and overall experience with Agentic GenAI applications is also required.
Experience with GenAI, LLMs, RAG, and prompt engineering.
Previous experience with at least one of the major LLM vendors (OpenAI, Anthropic, or Google) via their first-party SDK or a framework such as LangChain / LangGraph / CrewAI / AutoGen / PydanticAI / LiteLLM.
A good understanding of the internals, scaling, and limitations of Transformer models is necessary.
Experience with NLP libraries such as spaCy, NLTK, Hugging Face Transformers, and gensim is highly valuable.
Responsibilities
Develop zero-to-one concepts for expert systems
Systematically discover and test prompt engineering best practices
Optimize data sets for prompt development, model training and evaluation
Help create and maintain infrastructure required for efficient prompt development
Test and assess open-source solutions for LLM application development including orchestration frameworks, tool interfaces, solutions for context management, etc.
Develop automated techniques for the design and evaluation of AI agents
Analyze usage data to gauge the effectiveness of AI solutions and iteratively improve
Other
Master’s degree in CS/ML/DS or a bachelor's with equivalent experience.
soft skills to bridge technical and business perspectives.
Hybrid Work Model: We’ve adopted a flexible hybrid working environment (2-3 days a week in the office depending on the role) for our office-based roles while delivering a seamless experience that is digitally and physica
