You can send a comment under any collaboration tool, just start the comment by tagging your Virtual Contributor and start collaborating.
Bring the power of AI into your collaboration.
What is a Virtual Contributor?
An AI persona you can work with on Alkemio. Safely.
There are many AI Personas out there, each providing knowledge and capability. Be it ChatGPT, a legal AI or an AI assistant.
Virtual Contributors allow you to bring them safely into your collaboration space. Ready to share insights, answer questions, and enhance your productivity.
Allow others to access your knowledge and expertise
Scale your impact by creating your own Virtual Contributor.
As a freelancer, educator or a student, you can now bundle your knowledge and style. Others can invite your Virtual Contributor, so you can provide instant value within their specific context.
Create your own Virtual Contributor
1. Use Alkemio to create your AI Persona: upload docs and describe the interaction style.
2. Invite others to test and give feedback on your Virtual Contributor.
3. Publish and share your Virtual Contributor.
Already have an AI Persona you want to use? Simply connect it to Alkemio!
Join our Early Access program!
Become a Virtual Contributor to add value to your business.
Join now to get exclusive Early Access benefits.
Safety
Alkemio is built to be a safe digital foundation for collaboration.
Safety in how you collaborate: Who can join. What topics are worked on. How information is shared and visible. Clear roles and working groups.
Safety in the platform you collaborate on: Open-source. Your data won’t be sold or misused. Algorithms are only to help you make progress.
Alkemio is steward-owned, meaning our mission to help society collaborate is set in stone and safeguarded.
FAQ
About usage of Virtual Contributors (VCs)
Yes, of course that's possible! But for the Early Access Program, we encourage you to focus / start with one.
At the moment that is not possible. Agency will be added in the future so it can operate more independently. This will always be a setting, so you have the control to allow this or not allow it.
1. Currently, in each thread you can only interact with one Virtual Contributor. So if you have tagged 1 of your VCs in a discussion, you need to start a new one to interact with a second VC. At the moment, this is disabled because otherwise 2 or more Virtual Contributors can collaborate together and this could get out of control. When we are able to control this we will enable this on the platform!2. Your VC will not respond to a message that is already a reply to another message. Thread replies will be enabled soon.
At the moment the VC can read the information in the messages where it is tagged, and basic information about the Space it was tagged in. Later on, the VC will be able to read the entire context of the space and community (if you let it).
You can change the setting to public or private.When you choose public, space admins of other spaces can see and invite your VC to join their space. You always have to accept or decline the invitation. You can also rent it out to scale up your reach / business.When you choose private, only you can see your VC profile and add it to your space. When it is in a space, the community in that space can see it and collaborate with it. This means that when the space is public, everyone taking a look in the space can see your VC. You decide the levels of privacy and access.
Yes you can! All VCs that have the setting to be seen public can be invited to join your space. It is up to the owner of that VC to accept this invite.
About your data usage
At the moment, you can post plain text in any collaboration tool and upload text based PDFs, Excel files and Word files. We are working to implement more file types and extensions!
The VC combines the information you train it on with the knowledge of the Large Language Model it runs on. We are working to make sure there are correct quotes when the VC answers by referencing information provided by the creator.
No, your inputs or data will not be used to train the Large Language Model on which the VC is based on.
We keep finetuning how the Virtual Contributors work, and have built in checks and balances to avoid AI hallucinations as much as possible. We also make sure the answers are professional and appropriate. However, the output of the VCs depends on the knowledge given by the user, and it may not (in general) always give correct answers. We do our best to provide the source from where the answer was given, and keep finetuning the VC for a better experience as the technology improves.