Skip to content
seno3Public

About

Resources

Stars

0 stars

Watchers

0 watching

Forks

Latest commit

 

History

9 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

hera

Workplace accountability scoring platform. Analyzes SEC filings, court records, news, EEOC data, and employee reviews to rate companies on a 1–10 accountability scale.

Full Commit History!

Previous Repo had exposed .env files :(. Reach out if you want any specifics! image

Architecture

Python scrapers → Snowflake → Cortex Search → Cortex COMPLETE (Claude Sonnet) → Express API → React UI
privateclaudetest-main/
├── ingestion/       # Python data scrapers + Snowflake loader
├── backend/         # Express + TypeScript REST API
├── frontend/        # React + Vite + TypeScript SPA
└── snowflake/       # SQL setup scripts (tables, search service)

Prerequisites

  • Node.js 18+
  • Python 3.8+
  • Snowflake account with Cortex enabled
  • MongoDB Atlas cluster
  • API keys: NewsAPI, CourtListener, Plaid (sandbox), Resend, Nessie

Setup

1. Snowflake

Run these in your Snowflake worksheet in order:

-- Creates database, warehouse, and tables
snowflake/01_setup.sql

-- Creates Cortex Search service for hybrid vector + keyword search
snowflake/02_search_service.sql

2. Ingestion Pipeline

cd ingestion
cp .env.example .env  # fill in credentials
pip install -r requirements.txt
# Single company
python run.py --ticker TSLA

# Multiple companies
python run.py --tickers TSLA,UBER,MSFT

Environment variables (ingestion/.env):

Variable Description
SNOWFLAKE_ACCOUNT Snowflake account identifier
SNOWFLAKE_USER Snowflake username
SNOWFLAKE_PASSWORD Snowflake password
SNOWFLAKE_DATABASE Default: hera_db
SNOWFLAKE_SCHEMA Default: public
SNOWFLAKE_WAREHOUSE Default: hera_wh
NEWS_API_KEY NewsAPI key
COURTLISTENER_API_TOKEN CourtListener token
NESSIE_API_KEY Capital One Nessie API key
SEC_FILING_LIMIT Max SEC filings per ticker (default: 3)
COURTLISTENER_LIMIT Max court opinions per ticker (default: 5)
NEWS_LIMIT Max news articles per ticker (default: 10)
EEOC_LIMIT Max EEOC releases per ticker (default: 3)
REDDIT_POST_LIMIT Max Reddit posts per ticker (default: 5)
REDDIT_COMMENT_LIMIT Max comments per post (default: 3)
WIKIPEDIA_ENABLED Enable Wikipedia scraping (default: true)

3. Backend

cd backend
cp .env.example .env  # fill in credentials
npm install
npm run dev            # runs on http://localhost:3001

Environment variables (backend/.env):

Variable Description
SNOWFLAKE_ACCOUNT Snowflake account identifier
SNOWFLAKE_USER Snowflake username
SNOWFLAKE_PASSWORD Snowflake password
SNOWFLAKE_DATABASE Default: hera_db
SNOWFLAKE_SCHEMA Default: public
SNOWFLAKE_WAREHOUSE Default: hera_wh
PORT Server port (default: 3001)
PLAID_CLIENT_ID Plaid client ID (sandbox)
PLAID_SECRET Plaid secret key
PLAID_ENV Plaid environment (default: sandbox)
NESSIE_API_KEY Capital One Nessie API key
MONGODB_URI MongoDB Atlas connection string
EMAIL_HASH_SALT Salt for hashing employee emails
JWT_SECRET Secret for signing JWT tokens
DEMO_MODE_ENABLED Enable demo mode for testing (default: true)
RESEND_API_KEY Resend API key for verification emails
RESEND_FROM_EMAIL Sender address for verification emails

4. Frontend

cd frontend
npm install
npm run dev            # runs on http://localhost:5173

The Vite dev server proxies /api requests to localhost:3001.

Data Flow

  1. User enters a ticker in the UI (or runs the CLI)
  2. Backend spawns python run.py --ticker <TICKER> as a subprocess
  3. Scrapers pull from SEC EDGAR, CourtListener, NewsAPI, EEOC, Reddit, Glassdoor, Wikipedia, and Twitter
  4. Documents are deduplicated and loaded into the Snowflake raw_documents table
  5. Cortex Search indexes documents (hybrid vector + keyword)
  6. Cortex COMPLETE (Claude Sonnet) generates a structured accountability analysis
  7. Results are cached in company_analyses for 7 days
  8. Express API serves results to the React frontend

API Endpoints

Method Route Description
GET /api/companies/:ticker Get analysis for a company
POST /api/analyze/:ticker Trigger a new analysis
GET /api/analyze/:ticker/status Poll analysis progress
POST /api/portfolio/scan Scan multiple tickers at once
GET /api/companies/:ticker/alternatives Find higher-scoring alternatives
GET /api/resolve/:query Resolve company name to ticker
POST /api/reviews/auth/request-verification Send email verification code
POST /api/reviews/auth/verify Verify employee email
POST /api/reviews/submit Submit an employee review
GET /api/reviews/company/:ticker/aggregate Get aggregated employee reviews
GET /api/plaid/link-token Create Plaid Link session
POST /api/plaid/exchange-token Exchange Plaid public token
GET/POST /api/nessie/* Bank simulation (Capital One Nessie)
POST /api/community/action Log a user action
GET /api/community/activity Get recent community activity

Data Sources

Source Data Collected Authority Weight
SEC EDGAR 8-K, 10-K, Proxy filings Highest
CourtListener Court opinions on workplace issues Highest
EEOC Press releases on settlements High
NewsAPI Company news articles Medium
Reddit Employee sentiment and discussions Lower
Glassdoor Employee reviews (proxy) Lower
Wikipedia Company history and context Context
Twitter Social media mentions Lower

Key Features

  • Accountability scoring — 1–10 scale based on severity, response quality, transparency, and pattern analysis
  • Issue tracking — sexual harassment, discrimination, retaliation, pay gap, hostile environment, assault
  • Employee reviews — verified via email, anonymous (SHA-256 hashed), weighted by recency
  • Portfolio scanning — paste tickers or connect via Plaid to flag problematic holdings
  • Alternative recommendations — same industry, similar market cap, higher accountability scores
  • Banking simulation — Nessie integration for demo portfolio tracking

Tech Stack

Layer Technologies
Frontend React 18, TypeScript, Vite, Tailwind CSS, Radix UI, Recharts, React Router
Backend Node.js, Express, TypeScript, JWT, Mongoose
Data Snowflake (warehouse + Cortex AI), MongoDB Atlas (reviews)
Ingestion Python, BeautifulSoup, edgartools, pandas
Integrations Plaid, Resend, Capital One Nessie, NewsAPI, CourtListener

Scripts

Frontend

Script Command
Dev server npm run dev
Build npm run build
Preview build npm run preview

Backend

Script Command
Dev server npm run dev
Build npm run build
Production npm start

About

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages