EmployHER

Inspiration

A résumé can show what someone has done without making it clear what they should do next. For women entering tech, finding a useful community, recognizing the value of existing experience, and choosing a career direction can be just as important as finding job listings.

We built EmployHER to help turn “Where do I start?” into a concrete next step grounded in the work someone has already done.

## What it does

EmployHER helps early-career technologists turn résumé experience and career interests into actionable plans: projects to build, skills to practice, ways to explain their accomplishments, and communities to explore.

Our interactive demo shows what that guidance should look like. Users can explore career directions, inspect recommendations, save up to three priorities, check off tasks, and download a detailed plan. Each action includes a first work session, specific deliverables, and criteria for knowing when it is complete.

The demo also highlights women-led organizations—including Rewriting the Code, Women in Robotics, and AnitaB.org—with practical suggestions for getting involved.

The judge walkthrough uses curated recommendations. Our connected résumé-to-AI workflow is implemented but still undergoing end-to-end validation.

## How we built it

We built EmployHER with Next.js, React, and TypeScript. Gemini powers the structured résumé extraction and career recommendation adapters, while Zod validates their outputs.

The connected architecture uses Auth0 for authentication, Tiger Data PostgreSQL with pgvector for profile storage and relevant opportunity retrieval, and Drizzle for database access. Backboard supports optional persistent career context.

We used Playwright to check browser workflows, keyboard navigation, mobile layouts, and saved progress, alongside unit and database integration tests.

## Challenges we ran into

The hardest product challenge was making recommendations specific enough to act on. “Build a robotics project” is vague. A plan with a 20-command evaluation set, recorded failure cases, and a short demonstration gives someone a clear place to begin.

We also had to preserve the meaning of résumé evidence. A result measured under specific experimental conditions should not become a broad claim about performance, and a skill missing from a résumé should not automatically become a skill the person lacks.

Connecting external services introduced another challenge: Gemini returned schema errors, availability failures, and rate limits. These failures reinforced the need for clear error handling and a reliable, explicitly labeled demo.

## Accomplishments that we're proud of

We’re proud of turning career advice into concrete work plans with tasks, deliverables, examples, and completion criteria.

Our demo builds on an advanced student profile rather than suggesting beginner work simply because the user is early in their career. It connects research, robotics, and software experience to meaningful next steps.

We also built a complete interactive walkthrough with saved progress, downloadable plans, mobile support, and women-in-tech community recommendations—and verified database, embedding, and Backboard integrations separately from the demo.

## What we learned

Personalization requires more than inserting someone’s name into advice. It means understanding their existing experience and recommending a next step that builds on it.

We learned that useful guidance answers three questions: Why this action? What should I do first? What will I have when I finish?

We also learned to distinguish working interface demos, tested service integrations, and a fully verified user journey. Each provides valuable evidence, but they are different milestones.

## What's next for EmployHER

Our next priority is completing and validating the connected journey from authenticated résumé upload to Gemini-generated recommendations and saved career actions.

We want to improve provider reliability, expand reviewed opportunities and community resources, and make recommendations more responsive to users’ goals, available time, and progress.

We’ll also test EmployHER with students and early-career women to learn whether its guidance helps them choose— and complete—a meaningful next step.