Inspiration
SafeContractor was created to protect consumers from fraudulent contractors, a major issue in cities like Boston. Our platform verifies credentials and offers AI-generated summaries for quick, informed assessments. Some of our team members know individuals who’ve lost their life savings to dishonest contractors with little legal recourse.
What It Does
SafeContractor is a database where residents can search for and verify the credentials of contractors in Boston by compiling information from obscure public records. In addition to detailed contractor profiles, the platform automatically generates AI-driven summaries that offer users a concise overview of each contractor’s qualifications and background. This dual approach—rigorous verification and clear, AI-generated insights—helps reduce the risk of scams and empowers users to make informed decisions.
How We Built It
Our frontend was designed using Figma and implemented with ReactJS, emphasizing a clean, responsive, and user-friendly experience. On the backend, we leveraged SQL with SQLAlchemy as our ORM, and FastAPI. Additionally, we incorporated an AI summary feature that analyzes contractor data and generates brief, informative overviews for each profile, streamlining the decision-making process for users.
Challenges We Ran Into
Data science is a dirty business. A lot of the times, we found data that was not normalized, had "weird things". Accounting for this took the bulk of our time. Moreover we are dealing with massive amounts of data. The tables of our databases currently have over 1.5 million rows! This took a while to index. So much data is also slow, so optimizing it required careful design of SQL keys.
Accomplishments That We're Proud Of
- We extracted via web scraping around 70k contractor license records from https://www.mass.gov/how-to/check-a-home-improvement-contractor-registration. This was not in an accessible format and required serious ingenuity to extract.
- Our frontend UI are well polished and responsive. Our search feature is highly dynamic and elastic (you don't need to type the exact name)
- The platform not only offers a polished, responsive user interface but also enhances user confidence through automatically generated, clear summaries of each contractor’s credentials.
What We Learned
- Hackathons are no joke. This stuff was hard to pull off.
- We learned its not the destination of the final product but the friends we make along the way. #TeamBobTheBuilder.
What's Next for SafeContractor
Moving forward, we plan to expand SafeContractor beyond Boston to cover additional regions. Future enhancements include incorporating user reviews and ratings, further refining the AI summary algorithm for even richer insights, and exploring partnerships with local authorities to bolster contractor verification. Our goal is to continually evolve SafeContractor into a comprehensive, trusted resource that empowers consumers to make safer, more informed decisions.
Built With
- fastapi
- figma
- github
- github-jobs
- javascript
- mariadb
- openai
- python
- react
- sqlalchemy
Log in or sign up for Devpost to join the conversation.