Holly

See how Serif helped Holly transform hiring with intuitive UX and bold branding. A deep dive into design, strategy, and digital innovation.

Introduction

Holly is modernizing government hiring with AI. By automating job description optimization and approval workflows, Holly helps local governments streamline hiring, reduce bias, and improve efficiency. Their AI-powered platform ensures that job postings are clear, inclusive, and aligned with agency needs. Holly is pre-seed stage company.

My Role

I worked directly with the co-founders to define product strategy and design the full MVP workflow. I led product design, ensuring seamless user experiences, and collaborated with the agency engineering team for QA support. Additionally, I built Holly’s marketing site in Webflow, refined the product pitch, and developed a comprehensive brand identity, including a logo, style guide, and social assets.

AI-Powered Hiring for Government: The MVP

The Project

Government hiring processes are notoriously slow and inefficient, burdened by outdated systems and bureaucratic workflows. Agencies struggle to craft job descriptions that are both clear and inclusive, often leading to postings that fail to attract the right candidates. Additionally, the approval process for job postings is fragmented and inconsistent, requiring multiple layers of review that delay hiring decisions. When I joined the team, Holly had recently signed with a pilot partner and needed to build (and ship!) a streamlined, AI-powered solution without adding complexity for HR teams.

Solution

I worked closely with the co-founders to design an intuitive "job designer" tool that addressed these hiring challenges head-on. The tool ingests existing job descriptions, analyzes their content, and suggests AI-driven optimizations to improve clarity, inclusivity, and alignment with agency needs.

To kick off this process, we mapped out the current paper workflow as well as a proposed software solution in FigJam.

FigJam state diagram
Example of a workflow diagram in FigJam

We realized quickly that we needed multiple permission structures - it was important for our government customers that internal and external personnel could provide feedback on suggested changes during the approval process.

User permissions matrix study
User Permissions Matrix - we iterated on this with the pilot partner team, as well as with the engineering team.

A key focus of the design was usability—government agencies often work with rigid, legacy systems, so it was essential to create an interface that was both powerful and easy to navigate. By integrating feedback loops with the pilot partner and iterative improvements, the "job designer" became a tool that HR teams can rely on to make better hiring decisions with minimal friction.

FigJam sketch of wireframe
Example of (very early) ideation on the approval process
High fidelity mockup of approval process
Final mocks for the approval process
High fidelity mockup of approval process
In the review process, we learned that many users may provide feedback. By designating a "document owner", one person is in charge of gathering feedback, resolving issues and comments, and "publishing" the updated version after approval.
High fidelity mockup of change tracking
We learned that change tracking was a critical (and arduous) task for HR teams. A very low lift solution was for us to allow users to export a "clean" version and a "track changes" version to PDF or Word. This integrates well with their existing workflows - where teams will print off "track changes" versions from Word to collaborate on, providing value immediately.
High fidelity mockup of table view
Home view of all job descriptions within Holly, including high level metrics. When onboarding a new customer, Holly ingests all existing job descriptions and scores them based on compliance, clarity, and comparison to similar roles across other organizations.

Results

The MVP successfully launched with the pilot partner, receiving highly positive feedback for its ability to simplify job description creation and approvals. As a direct result of the MVP's success, Holly secured two additional pilot partners, validating the need for AI-driven hiring tools in government. Since the initial launch, I have continued to collaborate with Holly to refine workflows, enhance automation, and incorporate additional AI capabilities based on real-world usage and feedback.

The momentum generated by the MVP has positioned Holly as a promising solution for modernizing government hiring, with ongoing iterations ensuring continuous improvement and scalability.