SafeCall

• 

  Transcription and summary of the conversation, along with the audio level.

• 

  Example of appropriate ways to handle a situation given before each simulation.

• 

  Page where user can interact with caller.

Inspiration

We were inspired by ElevenLab's Text to Speech and Speech to Text features. We aimed to solve the problem of how one would practice a 911 call in a realistic environment

What it does

SafeCall is a simulation that places users in the role of a real-world police dispatcher. It is designed to represent the intensity of real 911 dispatchers through fast-paced scenarios. It simulates a real caller through Elevenlab’s voices and records the user’s response. It then gives a summary and feedback on the user’s performance. The goal behind the project is to train both operators and regular people to think fast and work efficiently in fast-paced, high-stress environments.

How we built it

Towards the beginning of our development process, we quickly divided work amongst ourselves depending on our differing skills. We then created a Figma design covering the different slides that we had wanted to include. Based on our Figma, we created a React basis for each of our pages. We utilized ElevenLab’s Text to Speech, Speech to Text, and various agents to create a conversational aspect to our simulation, then finalized the design by connecting each aspect using React.

Challenges we ran into

Throughout our development process, we encountered several issues. We initially planned to change the background behind the Webspatial frames to simulate the caller. However, the Apple Vision Pro does not allow for backgrounds to be changed without a separate app. We attempted to create our own app, but that would hide the Webspatial frames. Instead, we opted to revamp our project to simulate the operator's experience, since we could do so without changing the environment. Furthermore, we utilized WebSpatial, which meant that some of our members were unable to contribute to the React code towards the end of development. Despite this, we were proud of what we were able to create.

Accomplishments that we're proud of

Overall, we're very proud of our project. We initially started our project as strangers who had met for the first time. However, we were able to come together and put a great project together. We're proud of how we were able to utilize our resources, especially those of the sponsors like ElevenLab, and that we were able to achieve our different programs and goals together cleanly.

What we learned

We learned a lot throughout our project. Most importantly, we learned how to utilize ElevenLab’s Text to Speech, Speech to Text, and various agents and incorporate them into our React code efficiently.

What's next for SafeCall

In the future, we plan to add several features. Firstly, the inclusion of additional features like scores and automatic reports will allow users to master their ability to talk under high stress. Secondly, additional scenarios that we haven’t been able to implement yet, such as other common situations, to which police are commonly alerted. Lastly, once Webspatial can change the Vision Pro background, we plan to simulate real emergency scenarios so people can practice calling 911 in a realistic environment

Built With

• elevenlabs
• react