The business problem is to streamline the flow of information between payers, healthcare providers, and various other stakeholders to deliver the right insights to the right places at the right times, driving better outcomes for patients, reducing friction in the health system, and lowering costs at UnitedHealth Group.
Requirements
- 8+ years of experience in machine learning or data science: data analysis, algorithm design, model architecture specification, machine learning algorithm implementation
- 8+ years of coding experience with Python or R
- 2+ years of experience leading development in a data science focused team using deep learning architectures
- 2+ years of experience in machine learning model development, transformer architectures, data trend analysis
- 1+ years of production experience with cloud and big data technologies
- 1+ years of agile development experience
- Experience with SQL databases and query language
- Experience in AzureML, AWS, or cluster computing architectures
Responsibilities
- Initiate and drive novel method development in machine learning from first principles to deployed solutions
- Provide innovative modifications and extensions to existing deep learning and statistical learning architectures
- Perform independent statistical inquiry into new areas and identify growth points and areas of untapped information
- Partner with Engineering, Business and Technical Product to make and meet project commitments in an agile framework; rapidly delivering value to our customers via technology solutions
- Mentor junior data scientists to successfully deliver software and business value
- Responsible for continuous improvement for alerts, monitors and telemetry
- Communicate with stakeholders, subject matter experts, and colleagues to identify new means of approaching problems in the medical coding field
Other
- Legally authorized to work in the US without any restrictions
- Adhere to UnitedHealth Groups Telecommuter Policy
- Full-time employment
- Mid-Senior level
- Bachelor's degree or higher (not explicitly mentioned but implied)
