EVERLEAP — Enterprise SaaS
Improving hiring decision quality through AI-driven screening and behavioural insights.
CONTEXT
Hiring teams today operate in high-volume environments where resume quantity overwhelms evaluation capacity. Most Applicant Tracking Systems optimise job posting and orchestration workflows, but they do not meaningfully improve decision confidence during screening.
The stated problem: workflow inefficiency.
The actual problem: lack of trust in candidate evaluation.
Everleap was conceived to address this gap.
I led product thinking, discovery, UX direction, design systems and frontend architecture for the platform.
DISCOVERY & INSIGHTS
I conducted structured interviews with HR leaders across small and mid-sized organisations. Using JTBD mapping and behavioural analysis, we identified patterns in hiring workflows.
Key insights:
1. Screening fatigue reduces evaluation quality.
When HR reviews hundreds of resumes manually, decision accuracy declines.
2. Existing filtering tools optimise keywords, not competence.
HR teams reported low trust in automated ranking systems.
3. Hiring is risk mitigation, not speed optimisation.
Decision-makers prioritise confidence over efficiency.
The real opportunity was not automating more steps.
It was improving signal quality before human intervention.
REFRAMING THE PROBLEM
Initial product idea:
Automate job posting, resume matching and orchestration.
Post discovery reframing:
How might we improve hiring decision confidence before human screening begins?
This shift fundamentally changed the product architecture.
PRODUCT STRATEGY
We built an AI-powered hiring intelligence layer comprising:
• Custom resume scoring model
• AI-led structured first-round interview
• Comprehensive assessment summary
• Explainable scoring indicators
• Cross-role candidate memory system
Instead of replacing HR, the system supports structured decision-making. HR retains final authority.
The design principle was transparency over automation.
INFORMATION ARCHITECTURE
Hiring involves multiple cognitive layers:
Resume signal extraction
Skill verification
Behavioural indicators
Role-fit assessment
I structured the IA to separate raw inputs from interpreted outputs.
Dashboard design highlights:
• Clear score hierarchy
• Skill heatmaps
• Risk flags
• Contextual AI explanations
Information is layered progressively to reduce overload.
SYSTEMS & IMPLEMENTATION
I built an atomic design system to ensure scalability and velocity.
Key decisions:
• Component-driven architecture
• Storybook integration
• Standardised spacing and typography tokens
• Reusable evaluation modules
Frontend implementation aligned closely with backend AI structures to avoid interpretational drift between model output and UI representation.
This reduced ambiguity during development and accelerated iteration cycles.
LEADERSHIP & COLLABORATION
I led:
• Discovery sessions with HR stakeholders
• Vision alignment workshops
• UX direction and prototyping
• Design reviews with engineering
• Mentorship for junior designers
Worked closely with AI engineers to ensure scoring transparency and avoid black-box UX patterns.
Leadership focus was clarity and alignment.
OUTCOMES
• 4 enterprise MOUs signed pre-release
• Positive validation from HR leaders
• Reduced initial screening effort in pilot simulations
• Strong positioning as Hiring Intelligence Suite rather than ATS
REFLECTION
This project reinforced that AI systems must prioritise trust.
Automation without explainability creates resistance.
Designing AI requires balancing capability with human control.