VerifyHire

Inspiration

Finding real remote jobs online can be stressful, especially for students, newcomers, immigrants, and people trying to enter tech. Many fake job postings exist online, and it can be difficult to know which jobs are trustworthy. We wanted to build something that helps people save time, avoid scams, and feel more confident when applying for remote jobs. Since this hackathon focuses on breaking barriers in tech, we thought creating an AI-powered remote job verifier would fit perfectly.

What it does

VerifyHire is an AI-powered remote job verification platform. It collects remote job postings, analyzes them, and helps users understand whether a job looks trustworthy or risky. The platform can:

Fetch remote job listings from job APIs
Detect scam phrases and suspicious patterns
Check if a role is actually remote
Generate AI-powered legitimacy insights
Give trust scores for jobs
Filter and search opportunities
Allow users to save and track jobs they are interested in
Provide quick risk analysis before applying Our goal is to make remote job searching safer, easier, and more organized.

How we built it

We built VerifyHire using a full-stack approach. For the backend, we used FastAPI with Python to create API routes, scraping logic, verification systems, and database operations. We used:

FastAPI for API development
SQLAlchemy for database management
HTTPX for API requests and job fetching
OpenAI API (GPT-4o-mini) for AI legitimacy analysis
dotenv for environment configuration Our backend fetches remote jobs from RemoteOK, filters duplicates, checks language support, detects scam indicators, verifies remote validity, and produces legitimacy insights. For the frontend, we used Next.js + React to build an interactive user interface where users can: browse jobs search and filter listings sort by trust level track saved jobs manage application progress We connected the frontend and backend through REST APIs.

Challenges we ran into

One major challenge was handling job data quality. Different job postings contain inconsistent formats, missing fields, HTML content, and noisy descriptions. We had to clean descriptions, remove unnecessary formatting, and standardize data before analysis. Another challenge was balancing AI analysis with rule-based verification. We wanted AI insights, but we also needed fast and reliable scam detection using keyword logic and scoring methods. We also spent time debugging API connections, environment variables, frontend/backend communication, and deployment-related issues.

Accomplishments that we're proud of

We are proud that we built a working full-stack AI application in a short hackathon timeframe. Some achievements we are proud of:

Built a real backend + frontend system
Integrated AI analysis into job verification
Created legitimacy scoring and scam detection logic
Added remote validation checks
Implemented job tracking and saved opportunities
Built a clean searchable UI for users
Connected multiple technologies into one working product Most importantly, we created a project with a practical real-world use case.

What we learned

During this project we learned a lot about:
Building full-stack applications quickly under time pressure
Working with FastAPI + Next.js
Integrating AI APIs into production workflows
Cleaning and processing messy real-world data
Managing frontend/backend communication
Designing systems that combine AI reasoning with traditional validation logic
We also learned the importance of teamwork, debugging, and adapting quickly during a hackathon environment.

What's next for VerifyHire

We want to continue improving VerifyHire after the hackathon. Possible future features include:

More job sources beyond RemoteOK
Better scam detection models
Company reputation verification
Browser extension support
Personalized job recommendations
Resume matching against verified jobs
Application analytics dashboard
Multi-language support for global users Our long-term goal is to make VerifyHire a trusted assistant for safer remote job searching.