Co:there | Devpost

Landing Page
Coral's Response to: 인기 있는 음식이 뭐야? / What are some popular foods
Co:there's response to: 인기 있는 음식이 뭐야? / What are some popular foods
Coral's Response to: quels sont les aliments populaires? / What are the most popular foods around the world?
Co:there's response to: quels sont les aliments populaires? / What are the most popular foods around the world?

❗Inspiration❗

As we familiarized ourselves with each other, our team exchanged captivating stories of our travels, recounting unique experiences in diverse locations like Spain, New York, Banff, and several other places. The richness of cultural experiences and the diversity encountered during our travels inspired us to create an innovative application. Carrying on this excitement, we aimed to develop an application that 1) gives creative responses, 2) utilizes RAGS by providing real time data and 3) provides multilingual resource citations from the LLM

❓What it does❓

Co:there is a search engine powered by Cohere that empowers users to ask questions and receive responses with real-time information. It goes beyond by offering multilingual sources and identifying the language of the user's input prompt.

☁️ How we built it ☁️

We harnessed Cohere's APIs, including Co.Chat and detect_language, to establish baseline models and outputs. To fetch live data based on user input, we implemented a Python Selenium web scraper. The backend was developed using Django, while the frontend was crafted with HTML and CSS.

🏆 Accomplishments 🏆

It works!!! (And we did it in less than 5 hours)

✈️ What's next for Co:there ✈️

Multilingual Translation: Enable responses to be translated into other languages based on the user's input language.
Optimization: Speed up process of retrieving live data for RAG
Interactive Session: Transform the search engine into a chatbot, allowing users to engage in interactive sessions.
Performance Evaluation: Conduct statistical evaluations to assess and enhance the model's overall performance.