TiriVelo Admin Dashboard · Internal tooling, desktop first
What is TiriVelo?
TiriVelo is a Canadian pet care marketplace preparing to launch, connecting pet owners with service providers. The admin dashboard is being designed to support bookings, payments, support tickets, disputes, and day-to-day operations once the platform goes live.
Finish what the last team started, make it clean, make it modern, and get it to a place where engineering can build it without guessing. I want admins to open this thing and actually be able to do their job in it.
Michael Navarro · Founder and CEO, TiriVeloThe opportunity
As a team of four, we came together to shape TiriVelo's dashboard into a more cohesive and purposeful experience. We began with existing design files, work in progress, and the founder's initial vision, using these as a starting point to explore how the product could better support its users.
A key priority was helping admins manage everyday operations, from cancellations and no-shows to disputes, with greater clarity and consistency.
Our starting point wasn't a blank canvas. It was an opportunity to bring ideas together, establish a shared direction, and create a design foundation the team could continue to build on.
The problem
The TiriVelo admin dashboard had passed through three teams, with little continuity between them. There were no notes explaining past decisions, no consistent design system, and no clear record of why things had been built a certain way.
Before making the dashboard look cleaner, the team needed to make sense of what was already there.
Design process
Reviewed the current state, audited past teams' work, mapped screens to their flows, and recorded what was wrong or missing.
Asked whether testing had been done, or could be.
Split ownership across the team.
Reviewed with the lead and the founder.
Ready for engineering.
Taking stock
Before redesigning anything, the team needed a clear picture of what had already been completed and what was still missing.
Research
Since the founder mentioned that around 70% of the dashboard was already complete, the team requested usability testing before continuing with the remaining designs. The goal was to validate whether the existing direction was working, identify gaps, and check for accessibility issues before building further.
However, usability testing had not been conducted, and the plan was to test only once the dashboard was fully ready. The request was therefore not approved, so the team continued with the information available.
Accessibility became an early priority instead. The team ran the colors through contrast-checking tools, which confirmed accessibility issues. The findings helped prioritise the color system as one of the first areas to address.
Key decisions
Simplifying the color system
246 color values, many of them near-duplicates, and several failing contrast.
The palette was audited and reduced to 15 intentional, accessible colors, using WCAG AA contrast requirements as a baseline. This created a more consistent visual language while making important information, such as status and alerts, easier to distinguish.
The result was a simpler system that supported both the visual goal of a clean, modern dashboard and the practical needs of admins using it throughout the day.
Rebuilding the architecture
An old sitemap was buried in an earlier version of the file. Outdated, but a starting point.
It was audited against the current product and restructured to reflect how admins actually needed to navigate between tickets, bookings, providers, and other areas of the dashboard. This rebuilt information architecture became the foundation for the rest of the work. Instead of designing individual pages in isolation, each screen now had a clear place within the larger system.
The new architecture organised the dashboard into clear sections based on what admins needed to manage: User Management, Verification & Trust, Bookings, Payments & Finance, Tickets & Support, Analytics & Reporting, Platform Settings, HR, and Help & Documentation, with related tasks grouped under each.
Building a system, not just screens
The file looked complete at first, but many elements were only static overlays.
The screens were not connected into complete flows. The team built reusable components with clear states - default, hover, active, and disabled - and connected the screens into interactive prototypes. This made it easier to see how the dashboard would work from one step to the next.
A new main page was first set up with card placeholders to define the overall layout and structure. As the card components were created, these placeholders were replaced with reusable components to keep the page consistent. This included key sections such as the Admin Actions panel, Service Timeline, Documentation, and Ticket Alert information.
The work I owned
Once the foundation was in place, the focus shifted to the core admin task assigned to me: resolving tickets for booking issues. Six key workflows were mapped and designed.
The color system was simplified to make the dashboard more consistent and easier to scan. The primary brand color was reduced to TiriVelo Blue #3387A3, with clear status colors for different ticket states: green #51B173 for resolved, orange #FF963B for in progress, and red #D02E2E for urgent attention.
The filter options were also cleaned up, and dropdown filters were added at the top of the table so admins could quickly narrow down tickets by the information they needed.
Six workflows
Resolving situations where the provider failed to attend.
Handling the reverse situation, with different resolution rules.
Requires admin intervention, and can be resolved in three ways.
Managing a booking that was never confirmed.
Tracking a missing provider report and deciding what happens next.
Reviewing and resolving a cancelled service.
In practice
Each workflow was carried through to the screens an admin would actually use. Here is one of the six, end to end: a provider no-show, from the moment the strike lands to the two ways it can be resolved.
Step one
The owner reports the no-show and a strike is applied automatically. The provider gets seven days to appeal, and the deadline sits at the top of the admin panel. The Service Timeline is a collapsible card, so an admin can open the booking's full history or keep the view on recent activity.
Provider files an appeal
Step two
Filing an appeal pauses the strike and moves the booking into review. The provider's reason and evidence replace the countdown, and the admin gets two decisions - dismiss or apply - each behind a modal that requires a reason and a written note before it will save.
Two ways it resolves
Appeal rejected
Appeal approved
Outcome
The project moved from a collection of disconnected screens to a more coherent admin system, with:
Most importantly, the work addressed the original challenge behind the project: making the dashboard understandable and usable rather than simply making it look finished.
Reflection
The biggest challenge wasn't designing any one screen. It was designing within a project that had accumulated decisions without preserving the reasoning behind them.
A lot of the work involved figuring out what existed, what was still relevant, and what needed to change before new screens could even be designed. That experience reinforced how important documentation and consistency are in a product worked on by multiple teams.
Thank you for reading.
Happy to walk through the Figma file and the prototypes - get in touch any time.
Can't see the resume here? Open the PDF