ML Analyst for Neurojurist Service

Neurojurist is an AI assistant for lawyers that helps answer legal questions, analyzes documents considering context, and provides links to sources. It can advise on a wide range of legal topics, parse contracts, compare agreement versions, and identify edits and potential risks.

We are building a product that must meet professional standards of accuracy and argumentation. We work with a large team of legal experts, and it is critically important for us to deeply understand the current quality of the product and the direction of its development. You will improve existing approaches to quality assessment and create new, more scalable and reproducible solutions.

What tasks await you

Product quality assessment through expert evaluation You will develop and refine a methodology that will allow the ML team and the product team to identify growth points and extract insights from data. You will work closely with legal experts and directly influence the improvement of the model's quality.

Development of tools for data labeling Our products require a significant volume of textual data for training and validation. You will build a full cycle of labeling and preparation of training and validation datasets, organize quality control processes, and ensure the reliability and consistency of labeling.

Comparative analysis of quality for internal product configurations and external models Your task is to create tools for statistically correct comparison and benchmarking of solutions based on expert labeling and automated quality assessment methods, so that the team can make informed decisions about product development.

We expect that you

• Are proficient in SQL and Python
• Know mathematical statistics and probability theory
• Possess structured thinking and can formulate conclusions based on data
• Understand the basics of A/B testing and evaluating product changes
• Know the basics of ML

Will be a plus

• Have experience in ML
• Have worked with LLMs or agents