Soma-AI

Inspiration

I (Parth) personally love reading about philosophers and understanding the way they thought about life, death, and existence. I've always wished I could sit across from Plato, Nietzsche, or Simone de Beauvoir and have a real conversation about our views, to compare what they believed versus how I see the world today. Since that's not possible, I wanted to create the next best thing: a space where you can have personal, thoughtful conversations with the greatest minds in history, guided by your own feelings and ideas.

What it does

Sōma-AI is a personalized philosophical companion. When you start, you take a reflection quiz that helps match you with the philosophers whose thinking aligns most closely with your mindset. Once matched, the AI, powered by Gemini, trained only on authentic philosophical content, responds to your thoughts in the voice and style of your assigned philosopher. Sōma-AI also reads your mood through sentiment analysis and senses your environment through real-time weather data to adjust the tone of the conversation, making it feel truly alive and relevant to you.

How we built it

We built Sōma-AI using Streamlit for a clean, interactive frontend. The AI brain is powered by Gemini APIs, fine-tuned through context so that it generates responses rooted only in real philosophical data, not general chatbot replies. We designed a reflection quiz at the beginning to personalize the experience, pairing users with the right philosopher based on their answers. We layered in sentiment analysis to detect the user's mood and weather data to adjust emotional tone, ensuring the conversations feel as close to human, real, and present as possible.

Challenges we ran into

One of the biggest challenges was ensuring that each philosopher’s "voice" stayed authentic. It’s easy for a language model to sound generic, but it’s hard to make it reflect Plato's clarity, Nietzsche’s fire, or Beauvoir's subtle defiance. Another challenge was building the matching quiz logic, making sure the result wasn’t random, but actually meaningful. Finally, balancing technical complexity with a calm, inviting user experience, without overwhelming first-time users, took a lot of careful design.

Accomplishments that we're proud of

We’re proud that we didn't just build a chatbot, we created a living, breathing reflection tool where technology genuinely supports deep thinking. We’re proud of how well the reflection quiz, philosopher matching, sentiment adjustment, and weather personalization came together into a natural, thoughtful experience. We stayed true to our vision of making philosophy feel personal, not abstract, and that shows in every conversation users have with Sōma-AI.

What we learned

We learned that real personalization isn't about flashy features, it's about quiet, careful tuning to who the user is and how they feel. We learned how powerful it is to root AI in deep, meaningful data, in this case, centuries of philosophical thinking, rather than just making it "talkative." And we learned that the more human you try to make technology feel, the harder you have to work, but the reward is creating something that truly moves people.

What's next for Soma-AI

I want to make Sōma-AI accessible to all philosophy lovers - students, dreamers, wanderers, anyone who seeks to reflect deeper on life. The next steps are to open it up to real users, gather their feedback, and keep improving philosopher matching, conversation depth, and emotional intelligence. Ultimately, I want Sōma-AI to be a companion for anyone who seeks understanding, not just a bot. A place where thoughtfulness is honored, and reflection feels alive.

Built With

• gemini
• open-mateo
• sentiment-analysis-online
• streamlit