• Subscribe
  • I am building a AI chatbot publicly, and it is looking forward to your voice.

    Jenny
    3 replies
    Hi, friends, I am publicly building an AI chatbot to make the developer's documents better communicate with users. This construction idea comes from the problems encountered by our team: Our team has open-sourced multiple software, such as: mybatis-flex, agents-flex, aieditor. In addition to the daily work of maintaining the code, the most troublesome thing is that a large number of users ask us questions every day. Most of the questions are very basic, mainly because the documents are complete, but users are too lazy to find them. My team and I were deeply troubled by this matter, so we decided to use technical means to solve it. That is, develop a conversation robot, train it with git documents, and let it do this part of the tedious work instead of us. We decided to build this robot publicly, and now is the stage of drawing product pictures. The name of the robot is Gitmd.chat, and the domain name was registered yesterday. Usage scenario description: Use Gitmd.chat in the git page of my open source project AiEditor. When the user clicks [I want to consult], the conversation page will pop up. This robot is trained with git documents, so it will only answer questions about the content of AiEditor documents, focusing and professional. In addition to its highly targeted features, it can also help users quickly locate a certain position in the document, avoiding the situation where users prefer to ask questions manually because there are too many documents and it is like looking for a needle in a haystack to find the answer. In Gitmd.chat, you can choose a large model and use a private API key. We only provide basic software functions. This design has two purposes: You can control your consumption and recharge the large model directly, as much as you use; Your data will not pass through us, ensuring the privacy and security of your data. Set the pull frequency: Set the time according to the update frequency of your own documents: pull the document once every 30 days, 15 days, 7 days, 1 day, and 2 hours to ensure that the robot and the document are updated synchronously. Set the robot appearance: You can set the logo, color, welcome words, FAQs, etc. of the consultation page to be consistent with your own brand. The right side is the preview after setting. Set the robot menu: You can set the menu of the consultation page, which is located on the right side of the logo. You can divert users from the consultation page to related pages such as prices according to your own situation. After all the settings are completed, you can preview the conversation page, copy the generated code to the git page, and you're done in a few minutes! The above functions come from our own needs. I want to know whether the problems that Gitmd.chat can solve are common to developers. We are eager to hear your voice, whether it is good, bad, or even terrible. Let's work together to improve this product~~

    Replies

    heracleswellness
    Our process begins with a thorough inspection to identify the extent of the damage and the underlying causes. Whether it’s filling and sealing cracks, patching holes, or resurfacing the entire driveway, we execute each project with precision and care. At driveway repair Specialists, customer satisfaction is our priority. We strive to exceed expectations by delivering prompt service, transparent communication, and exceptional craftsmanship. Our goal is not only to fix your driveway but also to enhance its durability
    Gurkaran Singh
    Hey there! Building an AI chatbot to tackle those FAQs is like giving your codebase a personal assistant – smart move! Can't wait to see Gitmd.chat in action, making doc queries a breeze for users. Ready to revolutionize the way we navigate documentation? πŸ€–πŸ“š
    Share
    Jenny
    @thestarkster Hi~Singh, I am glad to see your words, which are really important to me. This requirement comes from our own team. We all feel that it is very useful, but we still want to hear from developers to ensure that this scenario is a common need of everyone. Your feedback is really important to us, thank you very much. We will also speed up the progress. Can we invite you to the internal test at that time?