Description
Introducing a prototype of a free and open source Kanban system that can run autonomous AI agents and assign tasks by a AI project manager.
That means that you can write an epic and then the AI PM takes over and splits it up in tasks for AI microworkers (or humans) to solve.
This has the possibility today to setup a task for anything a LLM can do in term of text analysis, researchers that can search and figure out facts for you, reading PDF, Word, Excel files, creating Google Maps, CSV files, scrape the web, generating images using any model, including your finetune, creating working questionaires, generating realistic audio conversations, generating charts, edit and cut in videos or use videos for generating web pages, any type of questions about any media and 1000s of other things.
The following services are already supported in workflow chains:
OpenAI GPT, OpenAI Dall-E, Elevenlabs, D-iD, SceneX, PromptPerfect, SerpAPI, Pixabay, ConvertBot, Deepgram, DreamStudio, ScrapingBot.
It also supports any of the 100k+ inference models on Hugging Face.
Soon generation of PDF, Word, Powerpoint, Webpages will be possible as well.
Watch the video for scenarios that can be solved and how it works, why it was built in Drupal and what are the next steps.
The different parts for Drupal:
Kanban system: https://www.drupal.org/project/minikanban
Agent System: https://www.drupal.org/project/minikanban_agent
Dynamical Agents: https://www.drupal.org/project/ai_interpolator_agent
To listen to the generated podcast:
https://podcast.workflows-of-ai.com/sites/default/files/privatepodcast.mp3
Chapters:
0:00:00 Introduction
0:03:19 Generate a working questionnaire
0:06:38 Research and generate a podcast on any topic
0:14:04 Download a video from Youtube and cut it based on your text instructions
0:15:56 Use custom finetune models to generate images
0:18:03 Clarifications and background to buidling it
0:31:43 How does it work
0:41:48 Setup a autonomous agent dynamically
0:57:49 What comes next
1:11:51 Contribute and for the Drupal community
That means that you can write an epic and then the AI PM takes over and splits it up in tasks for AI microworkers (or humans) to solve.
This has the possibility today to setup a task for anything a LLM can do in term of text analysis, researchers that can search and figure out facts for you, reading PDF, Word, Excel files, creating Google Maps, CSV files, scrape the web, generating images using any model, including your finetune, creating working questionaires, generating realistic audio conversations, generating charts, edit and cut in videos or use videos for generating web pages, any type of questions about any media and 1000s of other things.
The following services are already supported in workflow chains:
OpenAI GPT, OpenAI Dall-E, Elevenlabs, D-iD, SceneX, PromptPerfect, SerpAPI, Pixabay, ConvertBot, Deepgram, DreamStudio, ScrapingBot.
It also supports any of the 100k+ inference models on Hugging Face.
Soon generation of PDF, Word, Powerpoint, Webpages will be possible as well.
Watch the video for scenarios that can be solved and how it works, why it was built in Drupal and what are the next steps.
The different parts for Drupal:
Kanban system: https://www.drupal.org/project/minikanban
Agent System: https://www.drupal.org/project/minikanban_agent
Dynamical Agents: https://www.drupal.org/project/ai_interpolator_agent
To listen to the generated podcast:
https://podcast.workflows-of-ai.com/sites/default/files/privatepodcast.mp3
Chapters:
0:00:00 Introduction
0:03:19 Generate a working questionnaire
0:06:38 Research and generate a podcast on any topic
0:14:04 Download a video from Youtube and cut it based on your text instructions
0:15:56 Use custom finetune models to generate images
0:18:03 Clarifications and background to buidling it
0:31:43 How does it work
0:41:48 Setup a autonomous agent dynamically
0:57:49 What comes next
1:11:51 Contribute and for the Drupal community
Default Thumbnail
Medium Thumbnail
High Thumbnail
Standard Thumbnail
Max Resolution Thumbnail
Video ID
xpsFk3tzxwQ
Playlists
Channel