Hera.

Inspiration

We know people who have struggles to choose between two seemingly equal companies. They can not make the decision easily as they want to make sure that the choice they pick is best for their future in the profession. Most of the time, they lack important information about the companies such as the pay gap, workplace culture, and how women are treated when misconduct occurs. We wanted to fix that and so we built Hera. We want to make a change by creating a platform that can easily show differences between companies and could help in making a decision for the person.

What it does

Hera is an accountability platform that empowers professionals, especially women, to make informed career and financial decisions by revealing how companies actually respond to workplace misconduct. Instead of generic ESG scores, Hera provides a clear 1–10 Accountability Score based solely on a company’s actions regarding assault, discrimination, retaliation, and toxic culture. Key features: Company Accountability Reports: AI-powered analysis of public records (SEC filings, court cases, EEOC complaints, news, employee reviews) generates a detailed score showing whether companies investigate, take action, or retaliate. Portfolio Scanner: Connect your bank (Plaid) or Capital One account (Nessie API) to see accountability scores for every company you spend money at - so your dollars align with your values. Employee Verification System: Current/former employees can anonymously share their experiences through email-verified reviews, adding real insider perspectives without compromising privacy. Pattern Detection: Snowflake Cortex AI identifies whether incidents are isolated or systemic, and flags companies with repeated failures. Alternative Recommendations: For low-scoring companies, Hera suggests better alternatives in the same industry and market cap.

How we built it

We first used python scrapers to gather documents from SEC EDGAR, CourtListener, EEOC, NewsAPI, Reddit, Wikipedia, and Glassdoor. This data is ingested into Snowflake. Then we used Snowflake to analyze each company’s document corpus to generate scores, summaries, timelines, and pattern analyses. For backend, we used RESTful API in TypeScript for company data, reviews, and portfolio scanning, MongoDB for anonymous employee reviews with JWT authentication and Resend-based email verification, Plaid SDK for real bank linking, and Nessie API for sandbox simulation. For the frontend, we came up with the UI/UX designs in Figma. We worked with different visualizations such as buttons, search bars, and a task bar.

Challenges we ran into

We faced many issues with web-scraping in the beginning as EEOC and SEC scrapers frequently returned empty responses. We added Reddit as a source to gain insights from the community that isn't available in formal filings.
We first used SendGrid for authentication, but SendGrid repeatedly failed authentication. We had to switch to Resend and make it work.
We used Sandbox, which did not include investment holdings, and it kept breaking our portfolio scanner.
We needed to make sure that the website could collect reaLal employees voices while maintaining anonymity.

Accomplishments that we're proud of

We are proud of building a project that can have a real life impact and actually matters. We believe that this project when implemented can be pretty useful to the society. We also finally found the correct API keys, and integrated them into our project after a lot of trial and error. We are also proud that we were able to process dozens of documents in under a minute for a company.

What we learned

We learnt how to use Snowflake. We also used and integrated different APIs. We learnt more about working together in a team, solving merge conflicts, and communicating clearly.

What's next for Hera

Marketing and Outreach: We want to publicise this platform more, so that more women are encouraged to add their reviews, and there are more reviews for the women browsing for reliability purposes. Security: We want to make our platform more secure as we do not want data about the reviews to be compromised. Divesting tools: In the future, we want to add a “Divest” button to generate shareholder letters and find ethical ETFs.