A product with no design foundation
Screenloop is an early-stage startup building interview intelligence tooling, with ambitions to become a Talent Operations Platform. One designer, a small engineering team, and no design system, research practice or product strategy.
Lead design strategy, execution and nurture the team around it.
- Assess the brand, product and tech stack before designing anything
- Put design into team meetings and product decisions
- Plan and ship an incremental redesign for quality and usability
- Implement UX research and product discovery
- Build and document the Screenloop design system
- Connect product, engineering, customer success and sales
The design strategy
Three months with customers, engineers and the data. Talent teams were not struggling with software but with coordination. Not a recruiter tool, then, but a coordination layer for hiring.
One platform, one architecture
Interview Intelligence, scorecards, pipelines, reporting, scheduling, the ATS. Six surfaces, three product squads, two user types, one mental model rebuilt at every move.
Candidate Experience + CNPS
Measuring the recruitment process from the candidate's side rather than the hiring team's.
No integrated way to evaluate a recruitment process, or to drill into what candidates actually said about it.
A whitelabelled survey engine: custom questions and response types on desktop and mobile, CNPS dashboards, sentiment analysis.
Six products, one coherent system
The design system came first, a constraint rather than a side project. Within 18 months it was live, documented and referenced daily.
Three calls that shaped the outcome
Design system before delivery speed
It slowed us early, and it is why six products still feel like one.
Research as a product input, not a UX ritual
Every cycle tied to a decision on the table. No research theatre.
Shipped six-plus products, ATS and Talent Ops platform
Breadth was the strategy, not a side effect. One platform, not six bets.
Results that moved the business
400+ design and research tasks, 39% ATS-related. Usage and adoption from July 2023 to June 2024.
What this changed
Same decision, same system, translated into the language of the person who had to fund it.
Putting the design system first was right. Making the case in design language was not. Consistency and reuse were not problems a seed-stage company felt, and for two quarters design was why things shipped slower.
What funded it was the same argument reframed around a cost the CEO already paid: engineering time rebuilding the same patterns. I reached his language two quarters late. Now I open with it.