A pipeline workflow that makes ownership gaps and compounding bottlenecks visible, helping recruiters move from fragmented context to a single diagnostic view.
I led this project from research through prototyping, anchoring every decision in direct recruiter input. By centering the work on how accountability breaks down across hiring teams, the result surfaces not just what’s overdue, but who needs to act and where.
Context & Problem Space
Recruiters working urgent, high-volume requisitions operate under tight SLA expectations. The information they need to act is fragmented across stages, records, and follow-ups, and the cost of that fragmentation compounds hourly.
Project Goal
Help recruiters working a single high-volume requisition orient faster, spot bottlenecks earlier, and identify who owns the next action at each stage.
Problem Identification
What are the problems?
Existing tools flag what’s overdue, but not who needs to respond or why the delay is happening. Before acting, recruiters had to manually reconstruct the pipeline, scanning for overdue handoffs and tracing ownership, losing triage time to orientation.
What needs attention first?
Where is the bottleneck?
Why is this candidate slowed?
Before acting, recruiters had to piece the pipeline together themselves.
Under SLA pressure, this meant manually scanning for overdue handoffs and trying to determine who was responsible for the next step, thus losing time to orientation before any real triage could begin.
Every hour past the SLA window increases the risk of candidate ghosting and offer declines.
Response time and drop-off data. SLA enforcement is the primary lever and the data shows why.
Research
Understanding how recruiters work: five findings, three assumptions overturned.
Process
Semi-structured interviews with five internal recruiters across technical, operations, and cross-functional roles. Sessions lasted 20–35 minutes, focusing on daily workflow, SLA pain points, and accountability breakdowns. Every finding below emerged in at least four of five sessions.
Findings & Key Insights
1. Orientation consumed the time meant for action
Context was scattered. Instead of starting with a clear view of risk, recruiters reconstructed pipeline state themselves. Time lost to orientation was time unavailable for triage.
2. High-volume requisitions made urgency harder to interpret
With many active candidates, urgency wasn’t about one overdue record. Recruiters needed to understand how pressure was distributed to judge what needed attention first.
3. SLA signals lacked diagnostic depth
Overdue tags indicated something was wrong but didn’t explain severity, cause, or what kind of response was needed.
4. Bottlenecks formed at the stage level, not just the candidate level
Patterns of slowdown appeared around specific stages, approvals, or handoffs. Multiple candidates were affected by the same friction.
5. Ownership eroded as stakeholders multiplied
As work moved between recruiters, hiring managers, and compensation partners, responsibility for the next action became harder to track.
Assumptions challenged
I entered the project focused on three familiar solutions: better sorting, cleaner data, and stronger SLA visibility. Early interviews challenged each one. Recruiters didn’t need better versions of existing tools, they needed support for navigating a high-pressure workflow.
The deeper problem wasn’t tracking what was overdue, but clarifying who owned the next action while the clock was ticking. That shift moved the project from designing for SLA visibility to designing for SLA ownership.
A ranked list could support prioritization, but only after recruiters understood the broader distribution of candidates and pressure across the pipeline.
Existing tools centered on the candidate record; recruiters needed visibility into the stakeholder responsible for the next action. Better timestamps meant little if ownership was still hard to find.
Recruiters already knew when something was overdue. The pressing question was how to visualize who needed to act at each stage, enabling faster bottleneck diagnosis.
Assumptions challenged
I entered the project focused on three familiar solutions: better sorting, cleaner data, and stronger SLA visibility. Early interviews challenged each one. Recruiters didn’t need better versions of existing tools, they needed support for navigating a high-pressure workflow.
The deeper problem wasn’t tracking what was overdue, but clarifying who owned the next action while the clock was ticking. That shift moved the project from designing for SLA visibility to designing for SLA ownership.
A ranked list could support prioritization, but only after recruiters understood the broader distribution of candidates and pressure across the pipeline.
Existing tools centered on the candidate record; recruiters needed visibility into the stakeholder responsible for the next action. Better timestamps meant little if ownership was still hard to find.
Recruiters already knew when something was overdue. The pressing question was how to visualize who needed to act at each stage, enabling faster bottleneck diagnosis.
Reframe
How might we help recruiters under SLA pressure move from reading pipeline health to identifying who needs to act, without manual investigation?
Solution & Highlights
The case for the hybrid view built for both orientation and accountability
The workflow begins with orientation, then narrows into focused execution.
Recruiters first scan the Kanban to understand where pressure is building.
When recruiters identify a stage that needs attention, they move into the stage-specific workbench, where each stage operates as its own focused view with filtered candidates, acting owners, and the signals needed to act. Orient first, then execute at the stage level.
How might we surface accountability alongside pipeline urgency so recruiters can quickly identify who needs to act next?
Problem
The Kanban and workbench help recruiters see where pressure is building and where momentum is at risk. But they don’t fully reveal who owns the delay, whether the same stakeholder is driving repeated bottlenecks, or how widespread the issue is. That diagnostic step still depends on manual investigation.
Solution
A persistent right-side panel that keeps accountability data visible alongside the pipeline, helping recruiters move from spotting urgency to identifying who needs to act without breaking context.
What The Panel Adds
1. Ownership at a glance
The Accountability Summary shows which stakeholders are responsible for delayed work and how much of the backlog they own.
2. Pattern detection across candidates
Blocker Pattern Detection groups repeated delays under the same stakeholder, distinguishing isolated slips from broader slowdowns.
3. Severity and scope in one place
Overdue count, average delay, and impacted stages appear together with urgency and follow-up priority without switching views.
From orientation to action
The recruiter scans the pipeline and sees where pressure is building. A click on View Summary opens the SLA layer, shifting the view from what is happening to who needs to respond.
From pipeline to stage-level diagnosis
The SLA summary highlights the most impacted stage first. The recruiter moves directly into that stage, applies an overdue filter, and that filter persists across stages.
From overdue work to accountable action
The SLA panel layers urgency: counts and severity first, then impacted stages, then who owns the delays and where repeated bottlenecks emerge. Designed for diagnosis, not just detection.
Testing & Iteration
Exploring how SLA insight should live alongside the pipeline.
Process
Research pointed to a clear need: accountability data accessible without leaving the pipeline view. Three patterns were explored, each tested against the same criteria: does the recruiter maintain pipeline context while reviewing SLA information, and can they move from summary to action without reorienting?
Findings & Key Insights
1. Context loss was the biggest cost
The modal created focus but severed the connection to the board. Recruiters couldn’t cross-reference candidates while reviewing the summary.
2. Lightweight surfaces couldn’t hold the full diagnostic picture
The popover preserved some context but compressed accountability data to the point where blocker patterns, ownership, and severity signals couldn’t coexist.
3. The right pattern needed to support continuous reference, not momentary focus
Recruiters reference SLA data while scanning the board, clicking into stages, and deciding who to contact. The panel was the only pattern that supported that without interruption.
4. Discoverability is the tradeoff worth watching
The panel’s persistent but dismissable nature means it only works if the recruiter knows to open it. The urgency banner mitigates this, but trigger visibility remains the design’s most honest limitation.
Modal
The SLA summary takes over the screen as a centered overlay. It forces the recruiter into a dedicated review state, separating the summary from the pipeline entirely. Clear and focused, but at the cost of everything behind it.
Popover
The summary appears as a floating layer anchored near the trigger. It keeps the interaction close to where the recruiter initiated it and preserves partial visibility of the board. Lightweight, but limited in how much diagnostic depth it can hold.
Right panel (selected)
The SLA summary opens as a persistent side panel, keeping the Kanban visible and interactive alongside it. The recruiter can reference accountability data while scanning the board without switching modes. This is the pattern that best supports the orientation-first, diagnosis-second workflow the research called for.
Results & Impact
What the design changes.
Together, these changes shift Evereth from a candidate-by-candidate workflow to a more continuous diagnostic one: helping recruiters orient faster, spot ownership earlier, and act without losing pipeline context.
Triage compressed into a shorter, more focused workflow
Instead of moving between stages, opening individual records, and cross-referencing follow-ups, recruiters can scan the pipeline, open the SLA panel, and identify the ownership pattern without leaving the view.
Ownership became visible without opening individual records
The Acting Owner column and Accountability Summary made it easier to identify who was responsible for overdue candidates and where follow-up was most needed.
Stage-level diagnosis stayed focused as recruiters moved through the pipeline
Once recruiters narrowed in on a stage, the stage-specific workbench gave them a focused view for reviewing overdue candidates, scanning ownership patterns, and continuing triage without resetting their SLA lens at each step.
Accountability data stayed accessible without breaking pipeline context
The selected panel pattern let recruiters reference ownership, blocker patterns, and severity signals while keeping the pipeline visible and interactive. That preserved orientation during triage and supported a more continuous diagnostic workflow.