Developers who want to create their own chatbots are supported by Microsoft, which is deeply immersed in AI and Large Language Models (LLM).
Redmond announces Copilot Chat, a sample application based on the vendor’s open source Semantic Kernel Software Development Kit (SDK) that serves as the foundation for developing custom chatbots with natural language processing capabilities and speech recognition.
You can also train using your organization’s own data through an SDK available on Microsoft-owned GitHub.
“Our full sample takes advantage of advanced features such as multiple conversation topics, speech recognition, and file uploads to make your chats smarter and your bots smarter with your own updates. You get a persistent memory store that allows you to get smarter the more you use it, and you can even share it with others and participate in conversations with downloadable bots,” Microsoft’s Shannon Monroe wrote this week. increase.
Microsoft is actively working to integrate generative AI capabilities (mostly from OpenAI such as ChatGPT, GPT-4, DALL-E, etc.) across its product portfolio. The company unveiled his GPT-4-based Copilot at Microsoft 365 in March as an AI-enabled assistant to help businesses with various services like Outlook, Word, PowerPoint, Teams and Excel.
The company has since expanded its use of Copilot in other areas, including security products in March and Visual Basic in April.
Semantic Kernel Copilot Chat enables developers to create LLM-based chatbots for use in a variety of areas, from HR and customer service to e-commerce and educational environments.
“By providing personalized assistance and natural language processing, unique chatbots can improve the user experience for customers, students, and employees,” Monroe writes. “Users can quickly and easily get the information they need without navigating her complicated website or waiting for assistance from her customer service representatives.”
According to Monroe, such chatbots can help organizations improve efficiency (by offloading employees to more complex tasks), make personalized recommendations, improve accessibility, and increase employee recruitment. You can also meet growing demand without needing to do anything.
Developers interested in trying out Copilot Chat should update to the latest copy of the Semantic Kernel Github repository to ensure they have the latest version of the sample app, then go here for instructions and options to enable it. I need to check the functionality. Azure speech recognition and persistent memory store.
There is also a video showing the installation steps of the app. According to the supporting documentation, developers need Visual Studio Code, Git, .NET 6.0, Node.js, and Yarn to run Copilot Chat.
The app itself contains a conversation title, a conversation panel showing the various conversation threads the user is having with the chatbot, and conversation threads showing the chatbot’s responses and user prompts. There is also a box at the bottom of the screen where the user can enter a prompt. ®
