When our CEO challenged us to unify two diverged products into a single investigative platform, I led the design from synthesis to working prototype - and introduced a new way of building that the company is now adopting across teams.
*This work involves sensitive law-enforcement investigations, so I can't share full details publicly. These visuals are representative of the live product, and I'm glad to walk through the complete case study and design decisions in conversation.*
The Challenge
Design Question: How do you unify two diverged products when the inputs are coming from every direction at once?
The product challenge
LeadsOnline and LeadSuite had been growing apart, and investigators were feeling it: users kept asking LeadSuite for search capabilities that only existed in LeadsOnline. The product was built case-first, but investigators wanted the freedom to move across their caseload as it evolved. The CEO directive: stop patching the gap and unify them.
The process challenge
A redesign at this scale, with input from PM, engineering, and the CEO simultaneously, can't move through traditional Figma handoff. The real work wasn't generating ideas; it was synthesizing them into something coherent enough to build.
The Synthesis Role
My job was to hold all of it - PM explorations, CEO input, dev constraints, investigator workflows - and turn it into a coherent product. The result wasn't a Figma file; it was a running app that the entire LeadSuite team could react to.
"I didn't hand off a spec. I handed off a repo."
What the Prototype Revealed
The knowledge panel wasn't in any of the early concepts. It came from using the prototype and asking what would actually be useful to see alongside results. Investigators can already see the data; the prototype revealed they couldn't always interpret it quickly.
Rethinking the Person Profile
The earlier redesign introduced tabs to bring structure to what had been a wall of data. Unification expanded the data model enough to support a relationship graph - a visual way to explore connections between people, cases, phone numbers, and addresses. It turns the profile from a reference page into an investigative tool.
The Bigger Bets
Universal search
Over 25 years, LeadsOnline built more than ten specialized search forms. We're replacing them with a single AI-powered search, and the bet is that if it's genuinely more capable, the transition is worth making.
Lead ranking
User feedback on the current leads feature was positive, but every lead looked equally important. Surfacing more insights without a way to prioritize them just recreates the cognitive load we were trying to remove.
Case Builder
Case Builder is where the AI integration becomes tangible. Investigators upload sources - police reports, CDR records, victim statements - and the system surfaces potential suspects, NCIC matches, and confidence levels. From the same view, they can generate legal documents like property hold requests and search warrants. It turns case assembly from manual cross-referencing into an AI-assisted workflow.
Impact
Team alignment in one session
The full LeadSuite team reviewed the prototype as a working app, not screens or a deck. Stakeholders specifically noted that the running prototype changed how they could give feedback - reacting to real behavior instead of imagining it from static screens.
PMs requested the workshop
After seeing a ClearCase feature demoed in production code instead of Figma, the PM team asked for a workshop. I'm hosting it to teach them the new workflow. What started as a personal process is becoming a company practice.
Repo as handoff
Engineering is starting from my codebase, not a spec. The prototype is already opinionated about component structure, design system tokens, and interaction behavior - fewer clarifying questions, faster starts.
Stronger design system
Every prototype stress-tests the system. Components that hold up get promoted; gaps that surface get filled. The system improves because it's being used, not just maintained.
Reflection
Design and code aren't separate disciplines anymore
The cost of exploring in code is now close to the cost of exploring in Figma. That changes what a designer can own - and what they should.
Synthesis is a design skill
The hardest part of this project wasn't the UI. It was holding CEO input, PM explorations, dev constraints, and user research at the same time and making them coherent. That's product thinking, not just craft.
Prototyping reveals what planning misses
The knowledge panel came from using the prototype, not from a requirements doc. Some of the most important design decisions emerge from building something and seeing what's missing.