What are the responsibilities and job description for the Software Engineer - Machine Learning position at vRad (Virtual Radiologic)?
Local Twin Cities candidate preferred. We are note able to offer or accept sponsorship for work authorization at this time.
Summary:
The Software Engineer or Data Scientist will be focused on improving radiology workflows using large language models (LLMs). In this role, you'll work directly with radiology report text (and optionally speech), preparing data pipelines and fine-tuning models such as T5, BERT, Whisper, and other transformer-based architectures.
This is a hands-on engineering and research position — ideal for someone with experience in NLP, audio transcription, and model development.
Essential Duties and Responsibilities:
- Prepare and preprocess radiology report text for supervised LLM fine-tuning and evaluation
- Build and optimize data pipelines for training on structured/unstructured text (JSON, TXT, DICOM metadata, etc.)
- Work with speech-to-text models to transcribe and process radiologist dictation
- Contribute to model deployment workflows (Docker, Kubernetes, inference services)
- Review, analyze, evaluate, and translate business and user needs into software specifications
- Participate in estimating, planning, and scheduling software development tasks
- Participate in the design of software systems, user interfaces, web services, processing algorithms, communications methods, and other major components
- Develop software in C#, HTML, JavaScript, Python, and SQL using standard coding practices
- Perform unit, integration, and system level software testing
- Assist SQA with integration and system level software testing
- Participate in code reviews
- Provide technical support to project team members, business users, and management
- Analyze and perform performance tuning of software and database systems
- Lead medium-sized projects including detailed design, development effort, and timeline
- Mentor junior developers to aid in their development
- Design performance-oriented models, relational database structures, and access patterns
- Ability to participate in on-call rotations as needed to resolve support issues
- Other activities as directed by management
Qualifications/Education:
- Either:
- Bachelor’s degree in a related field and 5 years of directly related experience
- 7 years of directly related experience
- Experience with the following is required:
- Strong Python programming skills, especially in NLP / ML contexts
- Hands-on experience fine-tuning LLMs (e.g., T5, BERT, GPT, Whisper)
- Working knowledge of CUDA, GPU environments, and multi-GPU training
- Experience with real-time speech-to-text transcription
- Experience with the following is preferred:
- C#
- SQL (Microsoft SQL or other relational databases)
- Unit, integration tests and validation
- Microsoft Visual Studio / Microsoft Visual Studio Code
- Microsoft SQL Server Management Studio
- Kubernetes
- Strong communication skills, written and verbal
- Self-motivated and able to work on a team on independently defined tasks
- Experience with the full software development lifecycle
- Experience with enhancement of object-oriented software applications using relational databases
- Experience with software and database architectural/design techniques
- Experience with highly available production software applications
- Demonstrated ability to own and work autonomously on medium-sized projects
Job Classification:
Sedentary- Exerting up to 10 pounds of force occasionally and/or a negligible amount of force frequently. A sedentary job involves sitting most of the time.
