over 1 year ago
Prague, CzechiaMid Level
Responsibilities
- Develop and improve NLP models for medical summarization, credibility analysis, and contradiction detection.
- Continuously improve automatic transcription pipelines and automate internal processes using AI.
- Ensure model accuracy, process stability, and optimized infrastructure and monitoring.
- Communicate technical machine learning results effectively in a business context.
- Collaborate with full-stack engineers to integrate ML models into production.
- Promote best practices across machine learning research, implementation, tooling, and system design.
- Conceptualize and implement new NLP ideas from validation through production.
Requirements
- At least 3 years of experience in machine learning roles involving end-to-end project ownership.
- Strong NLP background; ASR experience is a plus.
- Advanced experience with PyTorch or other modeling toolkits.
- Good understanding of software engineering principles and deep learning techniques, particularly for NLP applications.
- Good English communication and writing skills.
- Ability to work independently and collaboratively.
- Experience with or interest in legal or medical NLP applications is preferred.
- Ability to meet periodically at the Prague office.
Benefits
- Independent contractor opportunity rather than an employment relationship.
- Candidates must currently reside in or be willing to relocate to Czechia or Slovakia and have legal authorization to work there.
- Periodic in-person collaboration at the Prague office is required.
- Contractors are responsible for their own taxes and legal compliance.
Categories
About Filevine
Filevine builds a SaaS legal work platform for law firms and in-house legal teams, combining case and practice management with document and contract management, e-signatures, timekeeping, billing, payments, and analytics. The company also offers AI capabilities (LOIS) to extract insights and automate workflows across matters. Founded in 2015 and headquartered in Salt Lake City, it is privately held and used by over 125,000 legal professionals.
