Scientific AI Engineer

We are seeking an autonomous Scientific AI Engineer who will work closely with the existing ImmGen computational team and immunologists to design and implement this agent. They will integrate ImmGen’s existing data resources and analytical capabilities with the AI-powered interface. They will not be expected to acquire or manage ImmGen data, nor rebuild analytical tools. Instead, they will determine how to expose, adapt, orchestrate, and extend those tools so they can be reliably accessed through natural-language interaction, using MCP strategies that can be exploited by AI agents. The position will: • Work closely with ImmGen scientists and data analysts to understand existing datasets, metadata, documentation, tools, APIs, and access mechanisms, adapt and wrap existing ImmGen data so they can be invoked reliably by LLM-based agent that calls on public analytical and visualization tools. • Collaborate with ImmGen scientists to define supported use cases, desired answer formats, and appropriate scientific language and graphics. • Map existing and potential researcher questions to the appropriate existing ImmGen resources and analytical workflows. • Develop the software interfaces, APIs, schemas, computational guardrails and orchestration logic required to connect the agent to ImmGen’s existing infrastructure. • Implement safeguards against hallucinated results, invalid comparisons, inappropriate statistical interpretations, and unsupported biological claims. • Ensure that the agent does not automatically initiate unduly expensive computes. • We are looking for a team player eager to participate in the exciting re-writing of scientific data propagation. Basic Qualifications: • Minimum of five years’ post-secondary education or relevant work experience. Additional Qualifications and Skills: • Advanced degree or equivalent professional experience in computational biology, bioinformatics, genomics, or a related field. • Strong Python programming skills and experience with scientific computing libraries such as pandas, NumPy, SciPy, scikit-learn, Scanpy, Bioconductor, or equivalent tools. • Strong software-engineering skills, including experience with APIs, data schemas, testing, version control, logging, documentation, and reproducible workflows. • Experience integrating software with existing scientific data platforms, analytical pipelines, web services, or computational tools. • Hands-on experience with LLM applications, including tool/function calling, retrieval-augmented generation, structured outputs, agent orchestration, or natural-language-to-SQL/code systems. • Term:  This is a 2-year term position from the date of hire, with the possibility of extension, contingent upon work performance and continued funding to support the position. • Standard Hours/Schedule: 35 hours per week. • Visa Sponsorship Information: Harvard University is unable to provide visa sponsorship for this position. • Pre-Employment Screening: Identity • Interview Format:  The interview process will include both virtual and in-person components (e.g., initial virtual screening followed by on-campus final interviews). • Staying Informed About Your Application: Due to the high volume of applications, we may not always be able to reach out right away, but you can track your status anytime through the Careers@Harvard portal. #LI-DK1 Work Format Details This position has been determined by school or unit leaders that all duties and responsibilities must be performed at a Harvard or Harvard-designated location. Certain visa types may limit work location. Individuals must meet work location sponsorship requirements prior to employment. Salary Grade and Ranges This position is salary grade level 057. Please visit   Harvard's Salary Ranges   t