Rebuilding the analytics dashboard from the ground up
Project
Analytics Dashboard
Role
Sole Designer (Research, UX, UI, Prototyping)
Over 80 indicators redesigned across 5 pages. One unified dashboard to replace a fragmented system and help recruiters make smarter, data-driven decisions.
This is one of the largest projects I've worked on. What's shown here is a curated selection. Feel free to reach out if you'd like a deeper walkthrough.
Scope of Work
The problem
Analytics is one of the most critical features for Kelio ATS clients. Recruiters rely on it daily to track hiring performance, allocate budgets, identify trends, and understand what's working and what isn't.
But the existing system had serious issues. The data wasn't always accurate. The indicators lacked consistency. And the interface felt outdated and hard to read. On top of that, the dashboard was fragmented: free users all shared one basic view, while paying clients each had a custom dashboard built to their specific requests years ago. There was no unified product, just a patchwork of solutions.
The previous dashboards. Fragmented views, outdated interface, and inconsistent data visualization across client tiers.
The approach
The goal was to replace everything with a single, unified dashboard. No more custom setups, no more fragmented views. We simplified the model: either you have access to analytics, or you don't.
I led the entire design process from initial analysis to final UI. I ran the key decision-making meetings with analysts, project managers, and developers to align on what to keep, what to merge, and what to cut. We consolidated over 80 indicators across 5 distinct pages, each restructured around what matters most to recruiters.
One of five dashboard pages. KPI summary, trend analysis, source breakdown, and conversion funnel in a single view.
Making data readable
The biggest design challenge was accessibility. Our users are recruiters, not data analysts. Every chart, every number, every layout decision was made with readability as the first priority.
At the same time, the data had to be accurate and explorable. Recruiters who want to dig deeper should be able to, without the interface getting in the way. The balance between simplicity on the surface and depth underneath guided every choice.
Detailed analysis with data tables, comparative views, and entity-level breakdowns
A complete visual system
Every single indicator was reviewed and redesigned from scratch. Chart types, labels, spacing, color usage. Nothing was carried over as-is from the old system. The goal was a consistent visual language across all 80+ indicators, so users can scan any page and instantly understand what they're looking at.
Line charts: trend comparison, single metric evolution, positive/negative variation, and multi-layer area chart.
Vertical bar charts: multi-metric grouped, comparative, stacked multi-color, simple with filter, and weekly distribution.
Horizontal bars and ranking components: comparative view with tabs, categorized ranking, and monthly breakdown with totals.
Table components: sortable data tables with status indicators, detailed listings with badges, and campaign overview with team assignments.
Specific components
Beyond standard charts, I designed a range of specialized components to handle unique data types: NPS scores with benchmarking, qualitative feedback with ratings, funnel conversion with drop-off rates, device breakdown, and percentage-based distributions.
Purpose-built components: candidate feedback scores, NPS benchmarking, conversion funnels, device breakdown, and distribution charts.
Five pages, five angles
Each dashboard page serves a distinct purpose, giving recruiters a focused view on a specific aspect of their hiring performance.
Outcome
The project has been in active development for over a year. The scale goes beyond design: the entire data acquisition layer is being rebuilt by the engineering team to ensure accuracy.
The new dashboard replaces a fragmented system with a single, unified product. It gives every recruiter access to the same powerful analytics, presented in a way that's easy to read and act on. This is the kind of project where design doesn't just improve the interface, it shapes how an entire product works.
