← Back to Resume
Workday VNDLY  ·  2023–Present

Contingent Sourcing Agent (CSA)

🔒 NDA — Visuals password protected
Human-AI Interaction Agentic UX Explainable AI Enterprise UX Usability Research Early Adopter Programme B2B
My Role
Lead UX Designer
Company
Workday VNDLY
Research
July–Sept 2025
Method
60-min moderated interviews

Visuals are password protected

This project is covered by an NDA. Enter the password to view screenshots and recordings.

Incorrect password. Please try again.

Video not loading? Request a walkthrough.

Embedding AI into enterprise vendor sourcing — and making it trustworthy

The Contingent Sourcing Agent (CSA) is a VNDLY capability that surfaces candidate recommendations to help PMOs, Managers, and MSPs source contingent talent faster and with greater confidence. This case study captures the design and research work heading into GA — giving product and engineering partners a clear, shared reference point for the decisions made and the direction ahead.

Two goals shaped the work throughout: first, giving users recommendations that are transparent enough to act on confidently — not just a score, but a rationale; and second, standardising the terminology and UI patterns across the recommendations experience, moving away from inherited HiredScore language toward something that feels native to VNDLY.

👥
The Programme Management Team Member
PMO Sourcer  ·  Enterprise VNDLY client  ·  Daily platform user
Responsible for filling contingent roles quickly and cost-effectively. Manages relationships with multiple staffing vendors simultaneously. Highly process-driven — needs to trust any AI recommendation before acting on it, and needs full auditability for stakeholder reporting. Sceptical of automation that removes their sense of control.

Four themes shaped the design direction

Finding 01
Strategic Forecasting Tool

Participants saw CSA as moving them beyond purely reactive sourcing. By surfacing available, qualified talent immediately, PMOs could make smarter upfront decisions — including whether to pursue contingent vs. full-time hiring.

Finding 02
Worker Redeployment Critical

100% of PMO participants identified worker redeployment as a high-value feature. Organisations had no easy way to track or redeploy previously engaged contractors — talent was disappearing "into the ether" with no visibility into work history.

Finding 03
Trust Built on Explainability

A strong need for transparency in AI-generated grades and recommendations was expressed. Users needed to understand why a recommendation was made before acting on it — preferring AI as a "co-pilot" rather than a passive black-box feature.

Finding 04
Data Cleanliness Risk

Participants were concerned the AI would be making judgements on stale, incomplete, and intangible data — noting that candidate profiles are only updated when submitting for a specific role, leaving data perpetually out of date.

Designing a trustworthy agentic workflow — co-pilot, not autopilot

A validated AI sourcing agent that users understand and trust

The CSA v1 concept test returned strong signal on the value of strategic forecasting, worker redeployment, and work history visibility. The research directly shaped the design direction for the production feature — ensuring transparency and user control were baked in from day one, not bolted on.

5
Enterprise clients in Early Adopter study
100%
PMO participants valued redeployment feature
✓
Explainability & co-pilot patterns validated