Streamline your hybrid workspace with WiggleBot AI. Save time and effortlessly manage complex bookings for desks, meeting rooms, and more, ensuring your team stays productive and organized without the hassle.
"Book me a standing desk this afternoon"
"When is Penelope next in?"
The smartest office management assistant, at your service.
Make complex bookings in seconds with the intelligence of WiggleBot AI.
WiggleBot AI makes managing your hybrid workspaces easy-peasy.
From short chats, WiggleBot quickly understands your booking needs, handles complex requests, and ensures your team always has the right space. From desks to meeting rooms and hybrid schedules, WiggleBot AI works around the clock to save you time and increase productivity.
Please reach out to hello@wiggledesk.com
WiggleBot is powered by a powerful LLM integrated into a custom data pipeline built internally at WiggleDesk from the ground up to serve AI assisted experiences. We have fine-tuned our own Large Language Models based on industry leading large open source models. We self host these alongside our existing tech stack on Google Cloud.
Many organisations creating AI experiences simply build a wrapper on top of another AI Chat provider, however based on our years of assisting enterprises roll out hybrid working, we believe that by going slower, being more deliberate and creating our own self hosted AI solutions, we can achieve more effective AI experiences whilst retaining control of the data that you entrust with us.
The WiggleBot infrastructure is hosted in our existing Google Cloud platform. For AI inference we currently use Google Cloud managed datacenters in London, to ensure compliance with UK GDPR.
We are proud to be ISO 27001 certified, illustrating our commitment to managing information security to the highest internationally recognised information security management processes.
It's really important to acknowledge that AI solutions based on Large Language Models are an emerging technology. Whilst this field is improving everyday, it is certainly true that an LLM won't always predict the logically correct next word when chatting with a user.
We don't allow our chatbot to make any Create/Update/Delete operations directly. Anytime that a database editing related action is to take place, the user is prompted with a hard-coded widget requesting the user to make an explicit choice. This prevents LLM hallucinations, however unexpected from having any impact on data integrity.
Our inference pipeline is designed such that any workplace innapropriate conversations are gently steered back to office booking related activity.
No, chat content doesn't leave our existing secure Google Cloud environment.
As we have finetuned and are hosting our own Large Language Models internally, we are not reliant on external 3rd party LLM inference services.
I would definitely recommend WiggleDesk to others as the simple set up makes it easy to use and it covers everything that we need. We have now been able to add another building to our set-up, which means that we can have one desk booking system with everything in one place that everyone can access.
