NVP Knowledge Hub search and filter view

NVP Knowledge Hub

A searchable food-as-medicine resource hub for patients, providers, and advocates.

About NVP

Nutrition Value Pathways is HBOM's food-as-medicine initiative connecting Medicaid patients to nutrition benefits.

What are FAM & ILOS?

Food as Medicine (FAM) covers nutrition programs; In Lieu of Services (ILOS) lets Michigan Medicaid pay for them, but what's covered varies by plan and region.

The Problem

nutrivp.org told visitors that FAM and ILOS benefits existed, then gave them nowhere to go: no library, no filter, no way back in.

The Solution

A searchable hub where a provider, a patient, and an advocate each filter down to exactly what applies to them, without wading through content built for someone else.

Role
Design Intern
Team
HBOM, University of Michigan
Timeline
May – Aug 2026
Tools
Figma, Webflow, Adobe Illustrator

01

Problem

FAM and ILOS benefits are real but fragmented: coverage varies by Medicaid plan and region, and existing resources are scattered across providers, PDFs, and one-off pages.

nutrivp.org was a public landing page with no library behind it.

Patients

What does my plan actually cover?

Providers

No shared place to send patients to learn more.

02

Research

I reviewed existing FAM and health resource sites to understand what works and what's missing.

Competitive audit of three existing food-as-medicine and health resource site homepages

What was missing

  • No user-type segmentation: one page for everyone.
  • Unclear click behavior: no cue before you click.
  • No way to ask: no request or suggest path.

What worked

  • Fewer, clearer filters win at small counts.
  • Type and length cues cut click guesswork.
  • A warm, non-clinical tone stays approachable.

03

Design decisions

I explored multiple directions and weighed the tradeoffs before building.

1. Filter placement

Pain point · Filters were easy to overlook, so people never narrowed to what applied to them.

I compared filters on the side, on top, and both. Side filters are familiar and scannable but easy to miss on a busy page; top filters are eye-catching but crowd as categories grow. After team feedback, I used both, top for visibility and side for depth.

Filter placement options compared: top, side, and both, with tradeoffs

2. Resource detail layout

Pain point · The detail page had to hold a lot without burying the next action.

I compared a banner-style detail against one with the form beside the resource. The banner holds a lot cleanly; the side form prompts sign-up but competes with the content and feels intrusive. I kept the banner detail and moved the form to the bottom of the page as one dual request and suggest form.

Resource detail page layout options compared, with tradeoffs

3. Card exploration

Pain point · Every card had to show type, credibility, and time without clutter.

I explored several card layouts and reviewed them with the team. We moved forward with option 5 and put the tags below the description so they don't overlap the image.

Several resource-card layout options explored and iterated with the team

04

Features

User type, topic, and media type filter panels

Non-gating filters

User-type entry + Topic/Media filters. "All" by default: non-gating, so overlapping users never have to self-label.

Combined looking-for and suggest-a-resource request form

One form, two purposes

One dual-purpose form (request or suggest); captures user type for future segmentation.

Resource detail page with provider logo, estimated-time indicator, and open-in-new-tab link

Built-in trust signals

Org logo/banner, estimated read time, last-updated date, and open-in-new-tab.

05

Final solution

NVP Knowledge Hub running in a browser

Next steps

Usability testing with patients and providers, a public launch, and growing the library past its first set of resources.

Takeaways

1
Health content needs trust signals

A logo, a source, an estimated read time, small cues that make an unfamiliar medical resource feel safe to click.

2
Designing for people who don't fit one label

The forgiving, non-gating filters were the direct payoff of realizing "who are you" doesn't have one clean answer.