A marketplace is only as good as what people can find. Halo had the demand and the researchers, but the two couldn’t find each other. The answer wasn’t a smarter feed—it was enough verified supply to make the platform worth returning to. As the primary designer on the supply side, I shaped the connected surfaces that turned every profile, listing, and institution into a stronger network.

Problem

An always-on feed is only as good as what's behind it. It has to keep surfacing something worth returning for, and that takes supply. Demand was proven, but verified supply wasn't. Discover could reach researchers beyond Halo's network, but they hadn't joined, verified their expertise, or signaled they were open to partnering. Without enough relevant, on-network supply, there was little reason for anyone to come back.

Solution

Build a flywheel. I shaped the connected supply-side surfaces so each one feeds the next: every new profile, listing, and institution makes Discover more useful, which brings industry back. On the other hand, this also gives innovators more reason to be on the platform knowing they have visibility to industry. It all builds toward Discover 2.0, the always-on destination for solution scouting, landscaping, and discovering emerging capabilities.

My role

I owned design across Halo's supply-side surfaces, reporting to the Head of Product & Design, at a 3:1 engineering-to-design ratio.

Owned · Partnering listings, institution and org pages, network search redesign, end to end
Prototyped · Discover 2.0, the destination those surfaces were building toward
Supported · PulseChecks

Network size

~15.7K

innovators in 2026, up from ~8,000 in 2024

Growth

+33%

network growth over a ten-month period

New profile claims

4x

claim-your-profile activity from the enrichment flow

What changed

5 surfaces designed, 4 shipped

Get found without lifting a finger

Researchers are discoverable from work they've already published, with no marketing and no upkeep.

The network keeps growing

Every new profile, listing, and page sharpens the next discovery, so value compounds as it grows.

The intelligence layer for R&D discovery

Discover surfaces proprietary, on-network research, not the public results anyone else can find.

Person holding a tablet displaying a website with a search interface titled 'Find solutions to your R&D needs, fast.' showing search results for precision agriculture including researchers' profiles and details.

The story

1

A flywheel that turns discovery into a reason to return

2

Three sides of the same flywheel

3

Five connected surfaces

4

What changed, and what's next

The strategic bet was bigger than better search. R&D industry experts currently use a patchwork of tools that produce one-shot and generic results. We wanted Halo to be always-on, something teams returned to between projects because it kept surfacing something new and relevant. That only works if there's enough supply behind the experience, so the surfaces were designed to feed a loop.

Verified supply

Real profiles, listings, and pages from innovators on the network.

Better discovery

More supply means more relevant matches to surface.

More demand

Useful discovery brings industry back, again and again.

More participation

Innovators see the interest, so more join, and supply grows.

Making Discover valuable meant solving three participation problems at once. **Personas are composites built from user interviews.*

Industry

Needs a reason to return

Speed to a solution
Passive discovery between projects
Relevant science over pitch

"I don't need a prettier pitch, I need to get to the right science fast. And half my job is keeping an eye on who's doing interesting work before I even have a project for it."

Ivan

Director of Open Innovation

Innovator

Needs a reason to be found

Visibility without self-promotion
Expertise that stays current
Make industry connections

“I’ve spent whole weekends on proposals and heard nothing back. Not a no, just nothing. Marketing myself isn’t the job, and it isn’t what I’m evaluated on.”

Roshi

Principal Investigator

University Admin

Needs a scalable way to represent supply

No marketing budget
Represent many researchers at once
Fewer manual leads

"I'm responsible for showing what hundreds of faculty are working on, with no marketing budget and a team too small to even keep track of it all."

Amy

Technology Transfer Office

The supply problem couldn't be solved with one feature. Over the last two years, I shaped five connected surfaces, each addressing a different part of the loop.

Screenshot of the Halo website homepage featuring a search bar to find R&D solutions and a list of featured innovators including Brenda Prager and Texas A&M University, with descriptions of their research focus and partnership options.

Screenshot of a PulseChecks platform interface showing natural antimicrobial solutions and startup collaborations, including messages between users Maya Patel and another person about fermentation and product optimization.

Results

What changed for each side

The same three sides, after the work.

Industry

Before:

Search scattered across generic tools and cold outreach.

After:

Discover surfaced relevant researchers, organizations, and opportunities already connected to the network.

"We were engaging with several other companies whose capabilities have become 100% obsolete. Halo is building on proprietary data, real ecosystems, real connection and creating real value. This capability is very unique. There is no overlap with anything else."

Dimitris Lykomitros

Innovation Collaboration & Partnerships, Cargill

Innovator

Before:

Invisible expertise, little time to market themselves

After:

Structured profiles and opportunities find them.

"I submitted a brief and ended up pitching to a company I'd never heard of. Now we're in the final stages of negotiating a quarter-million-dollar research project."

Paul Westerhoff

Principal Investigator, Arizona State University

University Admin

Before:

Manual effort to represent every researcher.

After:

Institutional supply could be built and maintained from existing public evidence. No maintenance required.

"This is exactly the resource that my partnership office needs to act as a navigator and matchmaker for research and industry."

Olusola Fasunwon

Director, Research and Innovation Services, University of Lethbridge

The supply-side work changed what Halo could be. Researchers who had been invisible became discoverable, industry teams could find solutions and experts beyond what conventional scouting surfaced, and every new piece of structured supply made the network more useful. Real partnerships are forming, with deals now in active negotiation.

That created the foundation for Discover 2.0: an always-on destination where search, organizations, people, and opportunities become one continuously improving discovery experience. The supply work came first, because Discover could only become valuable once there was enough worth discovering. I prototyped and tested that destination.

The goal was never just better search. It was discovery that kept working after the search was over.

Anyone can use tools like ChatGPT for scouting, but those tools take a half-hour just to get the prompt right, and even then they don't necessarily produce the right types of information. Discover on Halo learns what I'm looking for and continuously monitors on my behalf.

Halo | Overview

Making a research network worth joining

tl;dr

Summary

Overview

Halo connects industry R&D teams with researchers, startups, and technologies to partner with. Its AI-powered Discover feed was introduced to help them find the right solution fast, and to keep an eye on emerging technology between projects. It's personalized to each team and always up to date.

Problem

Solution

My role

Scale & Impact

~15.7K

+33%

network growth over a ten-month period

4x

claim-your-profile activity from the enrichment flow

5

surfaces designed, 4 shipped

Get found without lifting a finger

Researchers are discoverable from work they've already published, with no marketing and no upkeep.

The network keeps growing

Every new profile, listing, and page sharpens the next discovery, so value compounds as it grows.

The intelligence layer for R&D discovery

Discover surfaces proprietary, on-network research, not the public results anyone else can find.

The story

1

2

3

4

Strategy

Turning discovery into a reason to return

Verified supply

Better discovery

More demand

More participation

Audience

Three sides of the same flywheel

Making Discover valuable meant solving three participation problems at once. *Personas are composites built from user interviews.

Needs a reason to return

"I don't need a prettier pitch, I need to get to the right science fast. And half my job is keeping an eye on who's doing interesting work before I even have a project for it."

Ivan

Director of Open Innovation

Needs a reason to be found

“I’ve spent whole weekends on proposals and heard nothing back. Not a no, just nothing. Marketing myself isn’t the job, and it isn’t what I’m evaluated on.”

Roshi

Principal Investigator

Needs a scalable way to represent supply

"I'm responsible for showing what hundreds of faculty are working on, with no marketing budget and a team too small to even keep track of it all."

Amy

Technology Transfer Office

The work

Five connected surfaces built the flywheel

The supply problem couldn't be solved with one feature. Over the last two years, I shaped five connected surfaces, each addressing a different part of the loop.

01.

Partnering listings

Self-serve IP publishing in about three minutes, turning research opportunities into structured, searchable supply for Discover.

02.

Institution & org pages

AI-generated institution and org pages built from public research evidence, so researchers become discoverable without creating or maintaining a profile.

These researchers aren't LinkedIn users tending a profile. They're seasonal, project-based scientists, too busy to be return visitors. So we stopped asking them to onboard and started asking them to verify a profile already built from their public work.

03.

Network & organization search

From keyword matching to AI relevance— industry sees not just what matched, but why an innovator or organization fits the need.

04.

PulseChecks

Signal before commitment. A lightweight way to surface interested innovators without launching a full RFP.

05.

Discover 2.0

The always-on destination that brings network search, organization discovery, and partnering opportunities together. I prototyped and tested the experience receiving very positive feedback.

Discoverability isn't about expertise, it's about partnering strength. An institution can hold deep expertise and zero interest in partnering, a dead end for industry. So the pages surface what an organization is genuinely open to partnering on.

Results

What changed for each side

The same three sides, after the work.

Before:

Search scattered across generic tools and cold outreach.

After:

Discover surfaced relevant researchers, organizations, and opportunities already connected to the network.

"We were engaging with several other companies whose capabilities have become 100% obsolete. Halo is building on proprietary data, real ecosystems, real connection and creating real value. This capability is very unique. There is no overlap with anything else."

Dimitris Lykomitros

Innovation Collaboration & Partnerships, Cargill

Before:

Halo connects industry R&D teams with researchers, startups, and technologies to partner with.

Its AI-powered Discover feed was introduced to help them find the right solution fast, and to keep an eye on emerging technology between projects. It's personalized to each team and always up to date.

Making Discover valuable meant solving three participation problems at once. **Personas are composites built from user interviews.*