Oratori - Daily Speaking Challenges

Your adorable pocket companion for confidence and communication.

Comment

Inspiration

All of us on the team grew up with Tamagotchis, raising a little companion and watching it evolve was such a core memory. At the same time, we all struggled with public speaking while growing up. It made school presentations, interviews, and even just speaking up in class really stressful. So we thought, what if we could take something fun and nostalgic, and use it to help with something that’s always been hard for us? That’s how Oratori was born.

What it does

Oratori is basically a mix between a Tamagotchi and a speaking coach. You can talk to your companion casually and it remembers what you share, making practice feel personal and supportive. When you’re ready to train, you can practice speeches or challenges and get feedback on things like filler words, pacing, and confidence. Over time, you can track your progress and actually see yourself getting better, while your companion evolves alongside you, turning the hard work of public speaking practice into something fun, rewarding, and motivating.

How we built it

We built Oratori as a Unity application, where the companion and game interface come to life in a pixel-art style. On the backend, we used Python to handle the speech pipeline, converting the user’s speech to text, analyzing it for filler words, pacing, and volume, and then sending feedback. That analysis ties directly into the evolution system, so the companion responds dynamically and grows as the user improves. By blending Unity for the front-end experience with Python for the speech analysis pipeline, we were able to create an app that feels both fun and functional, bringing back the nostalgic Tamagotchi vibe while actually helping users build confidence.

Challenges we ran into

One of the biggest challenges was figuring out how to give feedback without discouraging people. We didn’t want it to feel like the app was punishing you, instead, it needed to feel supportive. Another tough part was tying everything together (speech analysis, the companion's persona, the game mechanics) into something that actually felt like a real app, all within hackathon time. An aspect we really struggled with was the LLM's and the speech to text conversion because it kept disregarding the filler words like "um" or "uh".

Accomplishments that we're proud of

We got a working prototype that actually analyzes speech and adjusts the difficulty of prompts. We’re also proud of the design direction, mixing retro pixel art with something as serious as public speaking ended up working way better than we expected. And most importantly, people who saw the demo said it felt like something they would actually use, which made us really proud.

What we learned

We learned how powerful it is to use gamification for something intimidating. Public speaking can be scary, but making it feel like a game makes it a lot easier to approach. We also learned that UX and design matter a ton, the way an app looks and feels can completely change how supportive it comes across. And on the technical side, we learned a lot about real-time speech analysis and how tricky it is to balance accuracy with encouragement.

What's next for Oratori

We see Oratori growing way beyond the hackathon version. In the future, we want to add customization so you can unlock skins and personalize your companion, special challenge modes like speed speeches or interviews to keep things exciting, and more advanced difficulty features like distractions, audience reactions, and stricter filler word limits. We’d also love to expand the memory system so the companion remembers what you’ve said in the past and tests you on it, making practice even more realistic. Long-term, we picture Oratori being available in multiple languages, helping students in classrooms, and even being used for workplace training to make public speaking less scary for everyone.

Built With

Share this project:

Updates