I turned a shelter's five separate spreadsheets into one service-tracking system, with grant reporting built into the work staff already do.
ProblemFive services, scattered spreadsheets
Staff tracked showers, meals, laundry, shelter and medication in separate sheets, so leadership couldn't show funders their impact.
What I didResearch to grant-ready workflow
Interviewed staff, mapped their services and designed the dashboard, client profile and grant report.
ResultAdopted and award-winning
Won Best Hack for Fourth & Hope at HackDavis 2025, with a projected 40% cut in admin work. The shelter reported using it afterward.
Role
Product Designer
Timeline
24 hours, April 2025
Client
Fourth & Hope
Team
Me + 3 developers
Methods
Stakeholder interview
Service mapping
Wireframing
Prototyping
The problem
The data existed, but it was split across five spreadsheets
Fourth & Hope, a shelter and service provider, logged its services across separate spreadsheets. Coordinating a shift meant checking several files, a client's history had no single home and leadership couldn't turn the records into outcomes a funder would read.
Showers
Meals
Laundry
Shelter
Medication
→
One systemLive queues, one profile per client and reports that build themselves from daily logs
What I did
From one staff interview to a working prototype in 24 hours
01
Interviewed staff and mapped every service
One focused interview showed what staff needed: log services fast mid-shift, keep each client's information together and turn daily work into evidence for funders.
Log it fastStaff record services while they work, so entry has to take seconds.
One home per clientHistory was hard to rebuild across sheets.
Proof for fundersLeadership needed outcomes, not just records.
02
Set the structure in low fidelity
Before any color, I mapped the three screens staff would use most: the dashboard, the client profile and the report. Select any frame to enlarge it.
1. Staff dashboard Live queues by service, plus a log that feeds the report2. Client profile Demographics, medication and service history in one place3. Grant report Metrics pulled from logs, with an AI-drafted narrative
03
Designed the staff dashboard and client profile
The developers built on these mid-fi screens while they set up the database: live queues and wait times on the dashboard, and one profile with demographics, medication and service history.
Staff dashboard: live queues and wait times for every serviceClient profile: demographics, medication and service history in one place
04
Generated grant reports with Gemini, straight from daily logs
I defined what a funder report needed to show: reach, outcomes, budget use, highlights and next-quarter goals. We then built Gemini into the workflow: the system sends a quarter of service metrics to Gemini, which drafts the grant narrative as a downloadable PDF. Staff review every draft before it goes to a funder.
Daily service logs→Quarterly metrics→Gemini drafts the narrative→Staff review→PDF report
Q1 grant report · sampleMock data from the prototype
487People served
3,120Meals
92%Bed occupancy
18Housing moves
14Completions
11Job placements
View highlights and next-quarter goals
Expanded the kitchen team to offer vegetarian meals
Four program graduates returned as volunteer mentors
Next quarter: pilot a job-training workshop and add digital intake forms
Key decisions
Show demand live, keep each client in one place, let reports build themselves
Live queues on the dashboard
WhyStaff needed to see current demand at a glance to run a shift.
TradeoffThe view is only as current as the last entry, so logging had to be fast.
One profile per client
WhyRebuilding a client's history across sheets was the biggest pain staff described.
TradeoffSensitive data in one place makes access control a must-have.
Reports built from daily logs
WhyStaff already log services all day, so the report shouldn't be extra work.
TradeoffThe report is only as good as the daily data.
Results
An award-winning prototype the shelter kept using
40%
less admin work projected once services log in one place
50%
better grant readiness projected, with reports built from live data
35%
faster documentation projected for staff during a shift
Adopted
Fourth & Hope reported using it after the hackathon
5 → 1
spreadsheets replaced by one real-time system
Winner
Best Hack for Fourth & Hope at HackDavis 2025
Percentages are projected outcomes of the prototype, not measured results.
What I'd do next
A cleaner, more accessible staff dashboard
Looking back with more experience, I'd tighten the visual system and make the dashboard faster to scan mid-shift. Here's the direction I'd take, built on the original structure.