As founding product designer, I helped enterprises reconcile the two numbers that never matched: what they bought, and what they actually used.
Zylo set out to create a category that did not exist yet: SaaS Management. Companies were spending millions on cloud software bought by dozens of teams, with no shared system of record for any of it.
I joined as the founding product designer. My job was to take an ambiguous, category-defining idea and shape it into an enterprise product that finance, IT, and procurement could trust with their budgets. There was no playbook to copy, so much of the work became defining the patterns that the SaaS Management space would later take for granted.
By the mid 2010s a typical enterprise ran hundreds of SaaS apps. Finance saw invoices. IT saw logins. Procurement saw contracts. Nobody saw the whole picture.
That blind spot was expensive in two directions at once. Overspend: paying for redundant tools, forgotten renewals, and licenses well beyond what the team needed. Underutilization: seats provisioned in bulk and then barely touched. Both were invisible until a renewal forced a panic.
The questions every admin wanted answered, and could not, looked like this:
Am I over or under deployed on this app?
Where are my unused and idle licenses?
How entrenched is this tool if I choose not to renew?
Is there a quarterly opportunity to cut or consolidate?
A representative enterprise account inside the product:
One vendor alone left 147 paid seats idle. Multiply that across 178 apps and the cost of not knowing becomes the product.
If we could not represent usage and spend in a way that both a CFO and an IT admin trusted, no dashboard on top of it would matter. So I started at the whiteboard, with the model.
I mapped every seat onto a frequency spectrum, from Not provisioned and Over-deployed (pure waste) through Quarterly, Monthly, Weekly and Daily (entrenched value). The two buckets on the left became the easy wins for reduction. This model became the backbone of how the product flagged waste, and the signature you see at the top of this page.
The truth lived in other systems. I worked through what each integration (Okta, OneLogin, Concur, NetSuite, Salesforce, Slack, O365) could actually give us: users and record source, licenses and consumption, sign-in location. In the same session I defined an action versus descriptive icon system, with a now versus future path, so the visual language could scale as features shipped.
Before I designed a final screen, I mapped the entire admin onboarding journey (codenamed Prokure internally). It traced the ideal path of full cooperation against the difficult path of missing inputs, across three phases.
The map made the risky moments visible early. The biggest was adoption of the browser extension that powered usage tracking. By naming that risk on the wall in phase one, the team could design for it on purpose instead of discovering it in market. Mapping the difficult path next to the ideal one is a habit I have kept on every product since.
A founding designer rarely gets it right on the first try. Here are three core surfaces, each shown as it matured. The throughline of every pass: push insight closer to the surface, and let the product scale without losing its way.
A clean overview, but mostly a report. It told you the numbers, not what to do next.
Added the working layer: redundant functions by spend, upcoming renewals, SaaS history and forecasting, and live sentiment.
Re-architected navigation into a persistent sidebar as features grew, so added depth never cost orientation.
Apps as a flat list with users, monthly sessions and renewal dates.
Added spend, accounts versus active accounts, category and third party ratings, so a buyer could judge value at a glance.
A supplier to application master and detail view, to manage spend at the vendor level, not just app by app.
Four connectors, most of them still marked coming soon.
A true marketplace organized by category, with live sync status on every connector card.
Per integration detail with provisioned versus active licenses and a daily data pull log, closing the loop back to utilization.
The finished surfaces answered the four questions directly, and turned a sprawl of invoices and logins into a place to make decisions.
Redundant functions ranked by app count and spend exposed overlap to cut. Upcoming renewals on a timeline removed the renewal panic. SaaS history with forecasting showed where spend was heading. Application sentiment surveys added the human signal behind the numbers.
A sortable table pairing annual spend with accounts versus active accounts, category, and outside ratings. The gap between accounts and active accounts is the underutilization story, made impossible to miss.
A marketplace to connect the systems where the truth already lived, and a detail view that reconciled provisioned licenses against active ones on a daily pull. This is the engine that turned the dashboard from a report into a live picture.
Zylo helped pioneer the SaaS Management Platform category that industry analysts went on to formalize. The framing I shaped here, usage measured as a spectrum and reconciled against spend, became a default way the category talks about value. More than any single screen, that is the part of the work I am proudest of.
In a data product, legibility is the product. The model and the language do more for trust than any amount of polish on top.
Three lessons have followed me from Zylo into every product since. Model the system before you design the screen, because a clear data model makes the interface almost design itself. Map the difficult path beside the ideal one, so the risky moment gets attention before launch, not after.
And in a category that does not exist yet, the most valuable thing a designer can ship is a shared language. The spectrum from waste to entrenched value outlived every layout I drew on top of it.