External Research Collaborator (Formal Mathematics & AI)
# External Research Collaborator (Formal Mathematics & AI) **RWS Group** · France · `Remote` · `Other` 💼 **Уровень роли:** `Senior` 🕒 **Статус:** *Опубликовано: сегодня* · *Источник: Himalayas (JSON API)* --- ### Top Skills & Match 🎯 **Ключевой стек роли:** `[AI-Researcher]` `[Research-Engineer]` `[Formal-Verification-Engineering]` `[Machine-Learning-Engineer]` `[AI-Research-Collaborator]` `[AI-Research-Associate]` `[Research-Collaborator]` `[AI-ML-Research-Scientist]` `[Senior-Machine-Learning-Mathematician]` `[AI-Research-Collaboration]` `[AI-Research-Scientist]` `[Mathematics-Researcher]` `[R&D]` --- ### About the Role Job Purpose Job Purpose We are seeking highly skilled External Research Collaborator to contribute to the development, scaling, and maintenance of one of our client´s open-source autoformalization ecosystem. This interdisciplinary role sits at the intersection of formal mathematics, interactive theorem proving (specifically Lean 4 and Mathlib), and generative AI engineering. Collaborators will focus on building, evaluating, and refining multi-agent pipelines that translate informal mathematical natural language into verified, compilable Lean code. Additionally, collaborators will play a critical role in expanding the mathematical coverage of our client´s dataset, ensuring mathematical faithfulness, proof integrity, and robust integration with various systems. Job Overview Key Responsibilities Autoformalization Pipeline & AI Tooling Engineering - Develop, scale, and maintain multi-agent AI pipelines designed to translate complex mathematical texts into verified Lean 4 code. - Evaluate pipeline performance, diagnose verification and compilation failures, and implement systematic engineering solutions to improve autoformalization reliability. - Research and implement automatic tactic generation and domain-specific proof search strategies to assist LLMs in formal verification. Dataset Curation & Verification - Expand and curate our client´s dataset by formalizing missing mathematical results, theorems, and proofs. - Conduct rigorous peer-reviews of AI-generated Lean statements and proofs to guarantee mathematical faithfulness, proof integrity, and logical validity. - Ensure optimal code quality and promote the idiomatic, modular reuse of the existing Mathlib library. 3. Community Engagement & Research Collaboration - Actively collaborate with the global mathematics, Lean, and Mathlib open-source communities to align formalization efforts and upstream reusable code. - Support domain-specific formalization projects (e.g., algebra, analysis, topology) depending on specialized mathematical background. - Develop robust evaluation methodologies to benchmark machine-learning-driven formal reasoning tools. Skills & Experience Minimum Qualifications: - Educational Background: Ph.D. or Master’s degree in Mathematics, Computer Science, or a closely related quantitative field with a strong focus on formal methods, mathematical logic, or theoretical computer science. - Mathematical Maturity: Exceptional mathematical foundation, with a proven ability to understand, translate, and verify graduate-level mathematical proofs. - Lean 4 Experience: Hands-on, practical experience writing formal proofs in Lean 4 and familiarity with the design and structure of Mathlib. - AI Tooling & Programming: Strong software engineering fundamentals in Python and experience working with LLMs, prompt engineering, and multi-agent developer tools. - Autonomy: Demonstrated ability to manage open-ended research and engineering projects independently in a remote or collaborative setting. Preferred Qualifications: - Active contributor to Mathlib or other formal proof repositories (e.g., Coq, Isabelle/HOL). - Background in Machine Learning for Code, Automatic Theorem Proving (ATP), or reinforcement learning for symbolic reasoning. - Familiarity with compiler design, AST manipulation, or parser development in the context of Lean. - A strong track record of open-source software contributions or research publications in formal methods, AI, or mathematics. 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 Values Get the 3Ps right – Partner, Pioneer, Progress – and we´ll Deliver together as RWS. Covid Vaccination - All RWS employees hired for positions that require working on-site at RWS offices, customer offices, travel on behalf of RWS, and/or in-person meetings will be required to comply with the RWS USA COVID-19 Vaccination and Testing Policy. RWS complies with federal, state, and local laws regarding accommodations related to this policy. Recruitment Agencies: RWS Holdings PLC does not accept agency resumes. Please do not forward any unsolicited resumes to any RWS employees. Any unsolicited resume received will be treated as the property of RWS and Terms & Conditions associated with the use of such resume will be considered null and void. Originally posted on Himalayas
- AI-Researcher
- Research-Engineer
- Formal-Verification-Engineering
- Machine-Learning-Engineer
- AI-Research-Collaborator
- AI-Research-Associate
- Research-Collaborator
- AI-ML-Research-Scientist
- Senior-Machine-Learning-Mathematician
- AI-Research-Collaboration
- AI-Research-Scientist
- Mathematics-Researcher
- R&D
Наблюдалась 2026-10-11, впервые 2026-10-10, источник — Himalayas (JSON API).