Our Technology Stack
Vercel and Next.js
User-friendly website interface and initial processing.
We built our website using Next.js, a modern web development framework, and hosted it on Vercel. This setup provided a fast, responsive, and always-available website where users could upload images and view their analysis results. Vercel also handled initial image processing and validation through serverless functions, ensuring that only appropriate images proceeded to analysis.
Groq
High-performance hardware for image analysis.
Groq provided powerful computing hardware optimized for high-speed, intensive computations. We used Groq's hardware to perform detailed image analysis based on the 14 FEATS criteria, ensuring users received their analysis results promptly.
Akash Network
Scalable and flexible computing infrastructure.
Akash Network hosted our backend services that required more control over the computing environment. By deploying our image analysis pipeline on Akash, we could scale our services up or down based on demand, ensuring reliability and cost-efficiency.
DataStax
Secure and scalable data storage.
DataStax served as our primary database for storing uploaded images, analysis results, and user data. Its distributed architecture ensured high availability and quick access to data whenever needed, keeping user information safe and accessible.
LlamaStack API
Advanced language model for generating explanations.
We used the LlamaStack API to generate clear and detailed explanations for each aspect of the image analysis. It provided advanced features like safety checks, multi-step reasoning, and context management, ensuring that users understood why each rating was given to their image.
How Our Image Processing System Works
Our system was designed to be simple and efficient, guiding images from upload to insightful analysis smoothly.
- Start by uploading your image on the Try It page of our website built with Next.js and hosted on Vercel.
- The image undergoes initial processing and validation by our serverless functions on Vercel.
- The validated image is then sent to our image analysis pipeline hosted on the Akash Network, where it is processed using Groq's high-performance hardware based on the 14 FEATS criteria.
- We store the image and analysis results securely in DataStax for easy access and retrieval.
- Our system uses the LlamaStack API to generate detailed explanations for each rating, ensuring clarity and understanding.
- Finally, we present the results in a clear report on our website for you to review.
Benefits of Our System
- Quick Results: Users received their analysis promptly without long waits.
- Easy to Use: Our website was designed for a smooth and straightforward experience.
- Clear Insights: Users could understand their analysis with detailed explanations.
- Reliable and Secure: Data was stored safely, and the system was built to be dependable.