About TLD Assistant

TLD Assistant is a proof-of-concept application that demonstrates the advanced capabilities of Large Language Models (LLMs) to interact with tools like WHOIS services for both generic Top-Level Domains (gTLDs) and country-code Top-Level Domains (ccTLDs). It showcases how LLMs can not only fetch data but also infer useful insights about domain names using these external services. This functionality, known as function calling, allows AI models to interact with various tools and APIs in real-time.

Important Notice: Demo Status

TLD Assistant is a demo application developed purely for research purposes as part of the National Core Project PN 23 38 01 01 at ICI Bucharest in 2023. As such, this tool is not guaranteed to be reliable nor is its availability ensured. The application is part of an ongoing experiment to benchmark the function calling capabilities of LLMs, and users should be aware that results may vary and should not be considered fully dependable for critical tasks.

For more reliable and consistent performance, we recommend checking the OpenAI GPT Plugin, which provides domain name expertise with more robust availability. You can find the plugin at this link: Domain Name Expert Plugin.

Purpose and Goals

The development of TLD Assistant was driven by the need to evaluate the capabilities of LLMs in real-world applications, particularly in querying and interacting with WHOIS services. This tool is intended to offer insights into how AI can streamline tasks related to domain names.

Core Features

  • SEO-Friendly Domain Name Generation: Offers creative and SEO-optimized domain name suggestions to help users find suitable names for their projects.
  • Domain Availability Check: Utilizes public WHOIS services for gTLDs and ccTLDs to check the availability of domain names.
  • Insightful WHOIS Data Analysis: Beyond simple queries, the assistant can infer meaningful information from WHOIS data, such as registration and expiration dates, and domain ownership details.
  • Domain-Specific Focus: TLD Assistant is limited to domain-related queries to ensure its accuracy within this focused area.

Technologies Used

  • SvelteKit: A fast, modern framework for building web applications.
  • Node.js: Server-side runtime environment for JavaScript.
  • TypeScript: Type-safe JavaScript for improved code reliability and scalability.
  • Langchain: A framework enabling LLMs to integrate with external tools, key for the function calling abilities of TLD Assistant.
  • OpenAI models: This application uses the GPT-4o-mini, which is a powerful quantized language model, optmized for chat-based interactions.
  • Ollama: Ollama is an open-source tool that allows users to run large language models (LLMs) like Llama 2 and other models, such as Mistral, directly on their own machines. The TLD Assistant application has been successfully tested with smaller models like Mistral and Llama 2.

Understanding Function Calling

The concept of function calling allows LLMs to interface with external services, APIs, and databases. This extends beyond simple question-answering, enabling the assistant to perform real-time tasks like querying WHOIS data.

To learn more about function calling, you can explore the following resources:

Acknowledgments

TLD Assistant was developed as part of the National Core Project PN 23 38 01 01 within the Complex System Engineering Department at ICI Bucharest. This project highlights the potential of LLMs for real-world applications, particularly in the field of domain management and analysis.