Inspiration
while working with ai tools i noticed many times output is not reliable and changes a lot. single model is not enough for important data validation. so idea was to use multiple llm together and make system more accurate and stable.
What it does
aicheckdata is ai powered data validation and scoring platform. it takes input data and checks using multiple llm models and gives final score and output. it helps users to trust ai results more.
How we built it
we used react for frontend and supabase for backend and database. multiple llm apis like gemini groq mistral are integrated with fallback system. we also used real time updates and session handling to show instant results.
Challenges we ran into
main challenge was handling different ai responses and making them consistent. also api limits and failures were issue so we had to build fallback logic. managing real time updates without breaking ui was also difficult.
Accomplishments that we're proud of
we built a working multi llm system with fallback and scoring logic. system is stable and handles failures properly. also created clean ui which shows real time progress and results.
What we learned
we learned ai is not always reliable so system should handle errors and not depend on single model. also learned importance of building fast and testing with real use cases.
What's next for Aicheckdate
next we want to improve scoring accuracy add more models and support more data types. also planning to make it scalable and add api so other apps can use it easily.
Log in or sign up for Devpost to join the conversation.