shypp
[ ← all work ]
// case study · QLU

QLU2: +70% User Sessions with AI Search

QLU Recruit is a search platform for executive recruiting firms.The first version was custom-built for Spencer Stuart. While powerful, it required heavy training and manual onboarding, which slowed adoption and scalability.

clientQLU
roleHead of Product Design
year2024–2025
outcome+70% user sessions
QLU2: +70% User Sessions with AI Search
// results
+70%
user sessions
// 01

About the client

QLU Recruit is a search platform for executive recruiting firms. The first version was custom-built for Spencer Stuart. While powerful, it required heavy training and manual onboarding, which slowed adoption and scalability.

// 02

The challenge

The original search UX was cluttered and complex.

  • 15+ filters crowded the left side
  • 3 different starting options: prompt, job description, filters
  • Multi-step flows with previews, popups, and manual edits

This caused:

  • Hours of manual onboarding
  • CS and Sales tied up in training instead of growth
  • Inconsistent, confusing results, and a loss of trust
  • Low adoption despite strong backend capabilities

A powerful tool that felt too complex to use, stalling growth.

// 03

The old screens

Where the product started: every entry point competing for attention at once.

QLU2: +70% User Sessions with AI Search screen
QLU2: +70% User Sessions with AI Search screen
QLU2: +70% User Sessions with AI Search screen
QLU2: +70% User Sessions with AI Search screen
QLU2: +70% User Sessions with AI Search screen
QLU2: +70% User Sessions with AI Search screen
// 04

The approach

As Head of Product Design, Ali led a full redesign of the search experience.

  1. Discovery and alignmentSurfaced pain points with CS, Sales and Engineering
  2. Defining successSpeed, adoption, training reduction
  3. Direction settingConversational AI as the default, filters kept for experts
  4. Iteration and reviewTested milestones, user feedback loops
  5. Stakeholder advocacyBeta data used to hold trust and alignment
// 05

The solution

A large, central AI-powered conversational prompt field serves as the default entry point.
A large, central AI-powered conversational prompt field serves as the default entry point.
The AI extracts filters automatically or asks clarifying questions. Users can preview, add, or remove filters manually or via prompt, keeping control …
The AI extracts filters automatically or asks clarifying questions. Users can preview, add, or remove filters manually or via prompt, keeping control and transparency in their hands.
The conversational AI panel stays visible on the left side of the results page. Users can continue refining searches there to get more relevant result…
The conversational AI panel stays visible on the left side of the results page. Users can continue refining searches there to get more relevant results without leaving the flow.
Expert users can expand the filters to adjust results manually. The chat field always remains visible, and clicking it collapses filters so the conver…
Expert users can expand the filters to adjust results manually. The chat field always remains visible, and clicking it collapses filters so the conversational area takes focus.
// 06

The impact

AreaBeforeAfterImpact
Avg. search time10–15 min5 min3x faster
Relevance & accuracyMixedConsistently high↑ Trust
Training time per userHours and long sessionsMinimalFreed CS/Sales time
Product adoption (LogRocket sessions)Baseline+70%Major lift
Learning curveSteepLowFaster onboarding
// work with us

have a product that needsmore than a designer?

one week from now, one flow of your product could be live and working.

hello@shypp.io