Inspiration

Are you tired of the stress and frustration that comes with reporting an accident to your insurance provider? At CollisionCare, we understand the challenges that drivers face when dealing with the aftermath of an accident. That's why we've developed a solution that simplifies the entire process and provides unparalleled convenience to our customers.

What it does

Our natural language processing technology accurately transcribes voice messages left by customers reporting an accident, if there is any information left out, we likeably prompt our user to enter the information. This information is then processed and shared with insurance companies to ensure that all necessary actions are taken promptly. Our platform also provides personalized recommendations for collision centers and eventually, towing services in the customer's area, allowing for a seamless and efficient resolution to the situation. We are also currently sending auto generated emails based on the info given by the customer to a collision center of the customers choice.

How we built it

Collision Care is powered by Flask, Respell, SQL, React, and Tailwind, we also incorporated Natural Language processing technologies to achieve our. We used a whisper model and a gpt4 model and created a "spell" which was the backbone of our technology.

Challenges we ran into

  • Converting audio to save it into an s3 bucket and then pull in into another API came with its challenges: delayed get requests, audio syncing, to name a few.
  • Manipulating data received from the speech to text was tricky as we had to format and comprehend the data being sent.

Accomplishments that we're proud of

  • Solving a real world problem seemed very doable with the help of AI!

What we learned

  • Further our knowledge in voice based NLP's and the challenges they come with
  • Organizing time and planning
  • SQL querying and data manipulating

What's next for CollisionCare

  • CollisionCare hopes to encourage a better customer - business relationship and aids in helping people in need in a timely manner.
  • In addition, we're continually working on incorporating location-based features that will provide even more helpful resources and recommendations for both customers and insurance companies.

Built With

Share this project:

Updates