University of California San Francisco
Office of
Collaborative Research

Funding & SupportAI Tools Open Call

Office of Collaborative ResearchRolling submissions

Tell us the tool you wish existed. We may build it.

Every lab has a task that eats hours: screening literature, cleaning metadata, drafting the same report every quarter. Propose an AI tool that would take that task off your plate. If it could help other researchers too and fits what our team does well, OCR puts its own engineering effort into building it with you.

Formally: OCR AI Tools for Research Request for Proposals

OCRFunded and run by the Office of Collaborative Research.

What you receive

Our team's time, from first sketch to a tool people use.

  1. 01

    Scoping

    We meet with you to sharpen the problem, check what tools already exist, and agree on what a useful first version looks like.

  2. 02

    Build effort

    OCR's developers and data scientists build the tool, with you as the first user and tester. Effort is set by OCR based on scope and capacity.

  3. 03

    Release to the community

    Finished tools are shared with UCSF researchers through OCR Research Resources, with credit to the person who proposed them.

What a good proposal looks like

Your problem first, and other people's too.

You don't need to know how to build it. We're looking for a clear, recurring problem in research work that AI could make faster, easier, or more reliable, and that more than one lab has.

Strong fit

Tools we're likely to take on

Problems that repeat across labs and can be solved with existing models and UCSF-approved infrastructure.

  • Literature and grant search, triage, or summarization
  • Extracting structured data from documents, protocols, or clinical notes
  • Cleaning, harmonizing, or annotating research metadata
  • Assistants for drafting routine reports, IRB materials, or progress updates
  • Interfaces that make an existing dataset or resource easier to query

Less likely

Better suited elsewhere

We'll tell you if an idea is outside our scope and point you to who can help.

  • Training new foundation models or large-scale compute projects
  • Tools that serve only a single study, with no wider use
  • Clinical decision tools intended for patient care
  • Projects needing access to data we can't use under UCSF policy

Eligibility

Who should apply

  • Anyone doing research at UCSF: faculty, staff, postdocs, students, and research teams
  • No technical background required; you bring the problem and the know-how of your field
  • Willingness to test early versions and give feedback

Priorities

We give priority to tools that

  • Solve a problem shared by many researchers, labs, or departments
  • Save meaningful time on repetitive or error-prone work
  • Can be built and maintained with OCR's current skills and infrastructure
  • Handle data safely and within UCSF AI and data security policies
  • Have a champion ready to test and help refine them

How to apply

From first conversation to launch.

  1. Step 01

    Propose

    Email a one-page proposal describing the problem and the tool you imagine. Sketches and examples welcome.

  2. Step 02

    Quarterly review

    OCR reviews proposals each quarter for community benefit, feasibility, and fit with our team.

  3. Step 03

    Scope together

    For promising ideas, we meet to define a first version. Others hear back with feedback or pointers.

  4. Step 04

    Build and share

    We build, you test. Once it works, we release it to the UCSF research community.

What to prepare

One page is enough. Plain language is better than technical detail.

  • 01The problem: what task takes too long or goes wrong, and how often
  • 02Who else has this problem: other labs, roles, or departments
  • 03What the tool would do, in a sentence or two, with a sketch if you have one
  • 04What data or documents it would work with, and their sensitivity (public, internal, PHI)
  • 05What success looks like: time saved, errors avoided, or new work made possible
  • 06Your name, role, and department, and whether you can serve as a test user

Review

How proposals are judged

OCR's Research Resources & AI Tools team reviews proposals each quarter and weighs:

  1. 01Community benefit: how many researchers would use it?
  2. 02Impact: how much time or effort would it save, or what new work would it enable?
  3. 03Feasibility: can it be built with current models, data access, and OCR capacity?
  4. 04Fit: is it within OCR's expertise and mission?
  5. 05Safety: can it meet UCSF data security, privacy, and AI policies?

If selected

What we ask of recipients

These opportunities build shared resources, so they come with a few commitments.

  • Serve as an early tester and give regular feedback during the build
  • Allow the tool to be shared with the wider UCSF research community
  • Acknowledge OCR when the tool contributes to published work

Questions

Common questions

No cash is awarded. What you receive is OCR's own design and engineering time.

Other opportunities

Feb 2026Biospecimen Pipeline SupportUp to a quarter of an OCR clinical research coordinator's time for a year, to launch or expand a clinical biospecimen stream.Learn more →1 per yearResearch Plan ReviewOne complimentary review a year of a research plan from an early-career investigator outside our partner programs.Learn more →Cohort 2Parnassus Community FundUp to $5,000 for events, programs, and projects that connect three or more Parnassus Institutes.Learn more →May 2026Hellman Clinical-Translational AwardUp to $100,000 for physician-scientists to embed with OCR or a CoLab and profile clinical samples.Learn more →RollingData Library Analysis SupportSupport to pursue new questions in the UCSF Data Library and build analysis tools others can use.Learn more →RollingSabbatical Assistance GrantUp to $25,000 to host a visiting scientist at UCSF, or to support your own sabbatical elsewhere.Learn more →

Before you apply

Talk through your idea with us.

We can tell you which opportunity fits, who to bring on as mentors or sponsors, and what makes a proposal competitive. A short email is enough to start.

Funding consult
Talk through your application
30 min
Free
You'll meet with
Vincent Chan, PhD
OCR Director and ImmunoX Chief Strategy Officer
Book a time with Vincent Ask about this opportunity
vincent.chan@ucsf.edu · or ocr@ucsf.edu