AI + Toolkits
- Andrew Zigler
Description
Function calling capabilities for Copilot. Support the freedom of choice for function calling “toolkits” in the same way we support freedom of choice for the LLM. Unlock more potential with AI by creating custom functions with a no-code, drag-and-drop visual tool like n8n.
AI + Vision + Toolkits + Themes: What if you could use Copilot to set your Mattermost theme? Mattermost thread & video
AI + Playbooks + Toolkits: What if Copilot could work with data from Playbook runs? Mattermost thread & video
How to be a no-code AI power user (Tuesday May 21st @ AIE)
The transformative power of AI can be harnessed by everyone, even non-programmers. This presentation delves into the practical applications of AI in no-code environments, leveraging tools like Mattermost, n8n, and ChatGPT to create powerful, automated workflows that are accessible to all. Using just drag-and-drop and chat, discover how to use AI to schedule meetings, write messages, prepare for calls and interact with your favorite task manager. As AI reshapes the landscape of software, understanding how to effectively integrate these technologies into daily workflows becomes crucial. Through live demonstrations of building and deploying AI-enhanced, drag-and-drop workflows, this talk provides a roadmap to becoming an AI power user.
Other resources
Mattermost threads
May 13th demo and explanation video from Nick to AZ
Video Conferencing, Web Conferencing, Webinars, Screen Sharing
Passcode: G.5@!YwcAI-generated summary:
Introduction to Function Calling: The AI Copilot in Mattermost leverages OpenAI's concept of tools or functions. These are predefined operations that ChatGPT can perform, such as searching the internet or sending a webhook.
Use of n8n: n8n is used as the automation platform in this example to orchestrate the function calls because it exposes its list of workflows/functions via an API call. This allows the user to build tools with a no-code, drag-and-drop tool.
User Interaction: The AI Copilot interacts with users by suggesting relevant tools or functions based on the conversation context. This interaction is similar to how ChatGPT offers various tools during its operation.
Conclusion: The demonstration emphasizes the ease and flexibility of integrating function calling toolkits into Mattermost's AI Copilot using n8n, enhancing the user experience by automating complex tasks through simple, predefined functions that can be built with a no-code/low-code visual editor.
Nick’s PR (experimental)
Enables “External Tools” in the AI Copilot configuration, allowing the usage of external tools like n8n to provide functions to the LLM
