SpecSyncAI | Devpost

Architechture
SpecSyncAI
Time and Money Spent building PC

Inspiration

Building a PC can be a daunting task, especially for beginners. It is believed that 30% of first-time PC builders reported experiencing regret after their builds. Many encountered compatibility issues, frequent crashes, and overwhelming choices, often leaving them frustrated and dissatisfied.

What it does

SpecSyncAI helps users confidently select and assemble PC parts by providing tailored recommendations based on user input and preferences, reducing the likelihood of regret and enhancing the overall building experience.

How we built it

We utilized Pinecone for fast semantic searches and the LLaMA model from NVIDIA’s Build Cloud to process user queries. We also sourced data from PCPartPicker to ensure we had the most comprehensive and accurate information on available components.

Challenges we ran into

Initially, integrating the various components of our system proved to be a significant hurdle. From gathering data effectively from various various sources to ensuring compatibility across different parts and getting the model to run on streamlit for the first time using streamlit, troubleshooting was very nice with the docs.

Accomplishments that we're proud of

Successfully created an intuitive platform that minimizes user regrets by providing reliable recommendations, significantly improving the PC building experience.

What we learned

The importance of user feedback became evident as we refined our algorithms and responses. Understanding user pain points helped us tailor our solution more effectively, leading to a more user-friendly interface. The amazing amount of things the AI workbench can do and how easy it is to configure it using the documentation