About
Intro
Most of my days go into B2B SaaS and HR tech, the tools that run payroll, hiring and people ops for companies that can’t afford to get it wrong. In that work the calls matter more than the screens: what to build first, what to leave out, and how much an AI system should be trusted to do on its own. AI can now draft a flow or a component set in seconds. Deciding which problem is worth solving, and where a system should stop, is still the job.
I’ve been the first design hire at a startup, building the practice up from nothing, and I’ve led inside teams that already had one running. Both taught me the same thing. Good enterprise design isn’t about looking polished. It’s about earning trust inside a workflow someone repeats every single day.
I started as the founding designer at an AI HR platform and built the design function from scratch, scaling to a team of 13+ designers and an experience used by 10M+ people across 60+ countries. Since then I’ve owned 20+ projects end to end, from design systems and analytics to HR workflows and AI assistants. The common thread is deciding what to build, what to cut, and how to prove it worked.
it's me
What goes into my work
1. Making the call, not just the screens
Anyone can generate a flow or a component set now. The harder part is deciding what a product needs, what it doesn’t, and what to build first. On Leena AI’s design system I kept the library to what the product required and built the table and filter patterns early. For ZenAdmin’s brand I held the palette to one blue. Each of those could have gone the other way.
calls over screens
grounded in real research, not guesses
2.Starting with why, not what
Before I touch wireframes, I want to understand what problem we’re actually solving. Sometimes the thing people ask for isn’t the thing they need. On Employee Engage, the support tickets looked like a service problem. Mapping them showed a role and information architecture problem, and that reshaped the whole redesign.
Designing for AI trust, not just AI features
Uncertainty is the hard part of agentic products. People need to know when to trust the AI, when to double-check it, and what to do when it’s wrong. At Peoplebox I set Nova’s autonomy by what an action touches, not by how confident the model is, so people could predict what it would do before it did it. Clear states, honest feedback and a plain way to recover are what let people stop babysitting the system.
confidence isn’t the same as correctness
systems outlast screens
Building the practice, not just the screens
At Leena AI I was the first design hire, so design systems, hiring and process were as much my job as pixels. Design work only scales if the practice around it does, so I work with engineers from the start, split ownership across designers and review for consistency.
What dive into my work
Timeline
Peoplebox
Lead Product Designer
Set up design SOPs and cross-functional workflows to cut inconsistencies across the platform. Led the redesign of the PDF reporting experience, doubling engagement, and the performance review workflow, cutting management time 20%. Designed Nova, the AI assistant, and the autonomy rules behind it, rolling it out in stages from the review form to goals, surveys, one-on-ones and agents. Review submission time fell 68%, and 6 of 10 reviews in the first quarter went through its AI draft.
Oct 2025 – Jul 2026
ZenAdmin
Lead Product Designer
Owned UX strategy end to end for the HRMS and employee productivity platforms serving enterprise customers. Set up the design function, built a design system from scratch, designed the performance management module from zero, and created the brand system. Usability testing cut task completion time 40%, and research lifted engagement 55%.
Mar 2024 – Oct 2025
Leena AI
Design Manager (Founding Designer)
First design hire. Built the design function from the ground up for an AI-powered HR SaaS platform that scaled to 10M+ users across 60+ countries. Built the enterprise design system that cut design delivery from 10 days to 4, redesigned Employee Engage to cut support tickets 72%, and took the assistant from a rule-based bot to an agentic colleague. Hired and mentored a team of 13+ designers.
Oct 2019 – Mar 2024
Techrefic Tech
Visual Designer
Led UX for 8+ product launches across SaaS and web platforms, from research through handoff, translating 40+ product briefs into user-centric solutions.
Apr 2017 – Jun 2019