Inspiration

We're tired of struggling during interviews, being caught off guard with a question, and not having enough practice ahead of time. So we created Hire Vision, an AI mock interviewer that's experienced through Augmented Reality.

What it does

Hire Vision is gonna ask you to upload your resume, its going to take your resume, send it to our GenAI and back down through our Natural Language Processing AI Agents, and then ask you humane questions about the information relevant to your resume. The greatest part is we do this all in an Augmented Reality Experience, using our VisionPro, yes, you can interview anywhere without having to look down at a phone or stop what you're doing. Just put your AR headset on and start practicing.

How we built it

We built Hire Vision with Cloudflare's hosting and implemented a Llama model that generated us questions based on the text from the resume that was uploaded as a PDF. Using FetchAI's web app Agentverse where we utilized Python to communicate between agents and set up a smooth natural language processing agent to make our questions sound as natural as any interview. From there we fetched it using an API request built in Swift that parsed the data into string, allowing us to play with the response as we want. Lastly, we displayed this all using Apple's RealityKit, creating a seamless AR experience

Challenges we ran into

The hardest part thus far was using FetchAI's Agentverse. Finding documentation on the uagents framework was tough because most pages led to 404's but after some time we found the right resources making it all click together. Once we figured that out, the rest of the process became clockwork.

Accomplishments that we're proud of

We've successfully learned 3 new concepts and applied them together in the same day! This for us has been a huge accomplishment, the dedication and hard work has truly given us a memorable moment.

What we learned

We learned so much! Learning to use FetchAI's AgentVerse was an experience alone from setting up Agents and connecting them with each other to simplify tasks creating a smoother process while communicating with our GenAI. Also, Cloudflare's amazing built in features to create GenAI thats hosted for us on their platform.

What's next for Undecided

Next, we want to implement this on your MetaQuest, Oculus, all AR devices, and then into your phones to be accessed anywhere and practiced where ever. We also plan to create a realistic 3-D human model that will be your actual interviewer, giving it that realistic feel of you being in an actual interview!

Built With

Share this project:

Updates