DexGem's LLMs

Our Large Language Models (LLMs) are sophisticated text-based models trained on extensive datasets, featuring billions of parameters.

While LLMs excel in various tasks, their ability to process real-time information, such as evaluating a new crypto project that emerged recently, may be limited. To address this, we augment LLMs with supplementary processes and technologies, ensuring they effectively analyze and deliver accurate insights.

Integrating Cutting-Edge AI Models

Our platform harnesses the power of advanced AI models, including GPT-4, GEMINI AI, MISTRAL AI, and GROK AI. These models analyze a comprehensive dataset stored in our database. The dataset is constantly updated with information retrieved from the internet about various crypto projects.

Data Processing and Evaluation

We categorize and evaluate crypto projects using these AI models based on predefined criteria and rules, refined by both AI and industry experts. Each model assesses the reliability of a project and assigns a score, ensuring a thorough and nuanced analysis.

Technical Implementation

Our LLM back-end processes are developed in C# and Python. We interact with the APIs of each AI provider, sending additional data extracted from our scrapers. These interactions occur in a secure cloud environment. Specific instructions, known as assistants, are provided to the models to ensure they perform the assigned tasks accurately, minimizing creative outputs and adhering strictly to the guidelines.

Real-Time Project Evaluation

The AI interaction software waits for new projects to be submitted for evaluation via internal HTTP requests. Alternatively, it selects unprocessed projects from our database. This approach ensures continuous and up-to-date analysis of new crypto projects.