Generative AI Graduate Fellow - Technical Research and Development
New Jersey State Office of Innovation
Software Engineering, IT, Data Science
New Brunswick, NJ, USA
USD 20-30 / hour
Scope of work
We are rapidly exploring the use of generative AI in government operations to improve the delivery of services and the experience of NJ residents. This exploration is via a pilot-based approach, wherein multiple pilot projects are rapidly developed, deployed, and monitored in order to understand how to use AI effectively and responsibly in government.
There are several projects at varying stages of the pilot pipeline: some are early in development, needing background research and engineering work to get off the ground. Some are more mature, with functional prototypes being integrated into agency operations, and now requiring development and execution of evaluation plans to measure efficacy, risk, and impact. Finally, we are developing resources to communicate learnings about generative AI use for operational and policy development purposes.
The Generative AI Graduate Fellow will work on multiple projects at different stages of development, contributing to research, engineering, evaluation, project management, user research, stakeholder communication, and written and verbal communication about the projects.
Potential Projects
Project one: Public comment analysis. LLMs are used via a webapp interface to analyze and organize public comments giving feedback to NJ government entities.
Project two: Interview and case summarization. LLMs are used to transcribe and summarize interviews as part of a critical agency process. Further information on cases is integrated into the LLM system so that users can easily find reliable information about a case.
- A tool to suggest content and design revisions for readability and accessibility Creating guides, demos, and communication around LLM use. The fellow may also contribute to guides targeted to state employees, with instructional material on using LLMs for tasks such as translation, data analysis, and content generation and revision. These guides and other work products will be communicated via demos, and through written and verbal means.
Continued development and evaluation of AI products. Two projects are in a mature prototyping stage. For each of these projects, some continued development to add final features and integrate into agency systems is necessary. However, the bulk of the work will be in testing and evaluation of these tools. The fellow will work with agency partners to develop and execute testing plans that assess accuracy, completeness, bias, agency workflow impact, and other elements of tool performance.
Prototyping LLM products. Several projects need earlier-stage development work, including
Experience, Skills, and Abilities
- Relevant education background: Currently pursuing or recently graduated from a graduate program in computer science, data science, engineering, and related disciplines. Strong applicants who are recently or are currently graduating with a bachelor's degree will also be considered.
- Demonstrated experience with developing, deploying, or evaluating digital tools involving large language models (LLMs), including through cloud-based APIs.
- Some experience working at all levels of the development stack, including but not limited to: frontend web applications, backend APIs and databases, and cloud infrastructure.
- Experience with technologies such as: python, modern JavaScript, React, CSS, HTML, JavaScript libraries and frameworks, Amazon Web Services (AWS), SQL and NoSQL document store persistence, and Git/GitHub version control.
- Research skills, including breaking down complex problems, creating custom evaluation plans, performing research interviews and other qualitative data collection, and data analysis for research outputs.
- Proven skill in tackling problems with curiosity, engaging in inquiry, and understanding user needs to develop optimal solutions.
- Dedicated to advancing the mission of the Office of Innovation and public service, with a strong commitment to accessibility and fostering a more inclusive society.
Compensation: Paid, hourly wage of $20-30 per hour dependent upon experience.