Summer 2026 AI/ML Informatics Intern

The Informatics function, within Information Technology Services (ITS), builds the data infrastructure, dashboards, and analytic tooling BPHC uses to identify and act on health inequities across Boston. The team partners across bureaus and programs - including the Bureau of Child, Adolescent, and Family Health (CAFH) - to standardize equity-disaggregated data, integrate public health dashboards, and incubate interdisciplinary innovation under Strategic Plan Pillar 3 (Data and Innovation).

The AI/ML Informatics Intern will design, build, and evaluate a prototype AI-assisted tool that solves a well-scoped problem inside BPHC. With the supervisor in the first two weeks, the intern will choose one of two tracks: (a) a conversational analytics chatbot that lets BPHC staff query approved internal datasets in plain language with cited, equity-disaggregated answers; or (b) a CAFH client-intake AI assistant that streamlines intake, screening, and resource referral for families seeking BPHC services. The work is grounded in BPHC's 2024-2027 Strategic Plan, Pillar 3, and centers responsible AI practice.

Specific duties include: (1) scope the project with supervisor and partner program staff; (2) inventory approved data sources with stewards and document sensitivity, equity-disaggregation availability, and usage constraints; (3) implement the AI/ML prototype on BPHC-approved infrastructure using open-source components, with citations and reproducible logs; (4) build automated tests for accuracy, citation correctness, and disparate performance across race/ethnicity, language, and neighborhood; (5) run user testing with 3-5 BPHC staff and incorporate feedback; (6) produce a deployment-readiness brief covering security, privacy, data governance, monitoring, and guardrails; (7) deliver a final cohort presentation. All work follows BPHC data-handling, privacy, and IT policies; PHI is excluded from any prototype unless explicitly approved by ITS, Legal, and the partner program.

Specific duties include: (1) scope the project with supervisor and partner program staff; (2) inventory approved data sources with stewards and document sensitivity, equity-disaggregation availability, and usage constraints; (3) implement the AI/ML prototype on BPHC-approved infrastructure using open-source components, with citations and reproducible logs; (4) build automated tests for accuracy, citation correctness, and disparate performance across race/ethnicity, language, and neighborhood; (5) run user testing with 3-5 BPHC staff and incorporate feedback; (6) produce a deployment-readiness brief covering security, privacy, data governance, monitoring, and guardrails; (7) deliver a final cohort presentation. All work follows BPHC data-handling, privacy, and IT policies; PHI is excluded from any prototype unless explicitly approved by ITS, Legal, and the partner program.