WFM Labs just installed our chatbot on their website. Ted Lango, the community's founder, contacted us less than a week ago, and we were able to create a unique one of a kind chatbot for Ted, based mostly on YouTube video transcripts, and put it into production in 3 days! 🥳
It's still got some glitches here and there, but already the chatbot should be performing quite well and provide high value to Ted's community members.
Ignoring the delivery speed we were able to have this time, mostly due to Ted turning around light greased lightning may I add, both the chatbot itself, and WFM Labs in general have some pretty unique traits.
Using AI to Decrease Attrition
First of all WFM Labs is an innovative Work Force Management community founded by Ted Lango having decades of experience with call centers. We all know all the talks about how AI is to replace us all in the near future as employees. Ted seems to look at this completely opposite. Ted and WFM Labs are obsessed with workforce retention, implying keeping employees around as long as possible. The reason for this is the high cost of replacing employees due to training, lost productivity due to losing knowledge as employees quit, etc. Hence the idea is that by scientifically approaching work force management in the call center space, while putting employees first, they can decrease attrition, resulting in higher profits for the company.
At WFM Labs Science and AI is actually saving jobs! 😁
Ted and his community are using some pretty impressingly innovative ideas, including ideas originally invented by John Von Neumann for the Manhatten project, in addition to having developed their own formula for calculating your work force management costs. Other things they're using to reduce attrition are simulation games, such as the Monte Carlo simulation and research into employee health.
Admittedly parts of WFM Lab's work is highly scientific, and goes a bit over my head, but we're super excited to have Ted and WFM Labs on board using our chatbot - And of course it doesn't really matter if we understand Ted's work, as long as our chatbot understands his work 😉
Creating OpenAI Context Data from YouTube videos
One of WFM Labs' unique problem was that most of their training data was in the form of knowledge from YouTube videos having the form of podcasts and interviews. You can find their YouTube channel here if you're interested in following them. The way we solved this together with Ted was by having Ted provide us with transcripts, for then to develop a custom import function, turning low quality transcripts, into highly factual small pieces of "information bursts".
The problem with YouTube transcripts is that YouTube videos and podcasts typically contains a lot of "welcome to" and "how's the weather where you live", etc. This adds zero value to an AI chatbot such as ours, so by intelligently using AI and specifically OpenAI, we were able to completely remove this "noise" from the transcripts, before using these transcripts as context data for the chatbot.
This allowed us to use Ted's YouTube transcripts as training material for his chatbot, without adding "noise" to the training material. You can see me explaining the process below.
Later we will create a detailed article about how to use YouTube and Podcast transcripts in your own chatbot, including the Hyperlambda code we created - But for now you can enjoy the above video where I'm explaining the rough idea of the process.
- Contact us if you've got YouTube or podcast transcripts you want to turn into an AI chatbot
A Community for Work Force Management
Ted is building WFM Labs primarily as a community, something you can see from how he's using wikis, interviews with other senior managers in the space on his YouTube channel, forums for question asking, etc. This allows call center managers to come together, share their knowledge with each other, while leveraging Ted's decades of experience in this space - Backed up by interviews with other industry titans, such as illustrated below where he's interviewing Bob Stella on his show.
However, why this is super exciting for us, is because WFM Labs is actually using AI to retain employees and not to get rid of employees. The irony here is of such an obvious character I feel it's justified to emphasize that point once more.
For WFM Labs the question is "how can we use AI to keep our employees longer"
And the above gives is great personal pleasure for obvious reasons 😊
Yet again, if you want to check out WFM Labs, or test their AINIRO chatbot, you can find it below.
Below is an example answer from WFM Labs' AI chatbot about maturity levels related to work force management.
Notice, the chatbot is not 100% perfect quite yet, and we will improve upon it as we get more data about its usage - But it should be good enough to answer most questions our readers can phrase related to the subject. So if you ever wanted to use AI to answer questions about how you can retain your employees in your call centers, feel free to visit WFM Labs and have a go with their AI chatbot 😊