My name is Cristian, I'm a software engineer and online coach and and I want to talk you through the first AI-native coaching application, Coach OS.
00:09
So if I go into the coaches side here you can see that I have six clients set up.
00:13
These clients are presented to me in order of who needs the most urgent attention and we can see a summary of some insights that coach OS pulled out of the data that they tracked.
00:23
If I click into Sarah here, you can see how she can simply track using voice notes and give us lots of unstructured input like this.
00:30
We can see that the AI analyses this unstructured input alongside alongside the structured inputs such as training, training performance, weigh-ins, steps, and so on.
00:39
And it looks for patterns and it will provide us insights based on these patterns, such as this mention of knee discomfort that's come up a few times and it will provide us with recommended actions acting as a really, really handy coaching copilot.
00:54
This is particularly useful because I no longer have to keep all of this unstructured context in my memory.
00:59
And on the structured context side, I don't have to be a data analyst to be able to assess whether they're progressing at the target rate.
01:07
AI will do all of this for me and simply present me with actionable insights where I can make the final call.
01:13
We can also see positive trends detected here.
01:16
We click into here to show a few more examples where the client has mentioned they are stressed due to work, and they're missing training sessions.
01:24
We can see that Coach OS may recommend that we make some sort of intervention.
01:28
If we go back to my point around structured data, we can see that Coach OS conveniently presents trends in how the structured data moves and we can see whether our performance in line with our targets and how much we're off by.
01:39
And then we can easily know what interventions we might need to take for a client.
01:44
So once we have all this new information and we have an idea of what we do next, this brings me on to my next feature, which which is the ability to coach at, the speed of thought.
01:52
When I'm presented with new information, whether I have recommended action or not, I have an idea of what the training programme adjustment will look like or or what a new client's training programme might look like based on our initial consultation.
02:05
The majority of my time is actually spent on building out this training programme or making these adjustments in an app or a spreadsheet.
02:11
However, now we can do this using natural language.
02:13
We can simply tell Coach OS what adjustment we want to make with text, or we can do it by recording a voice note and then we can send this off.
02:23
It takes a couple of seconds for Coach OS to make those adjustments.
02:26
We're then presented with an updated training programme immediately where we can make any manual changes that we want from that, or we can request further changes using natural language again, and we keep going through a loop until we have the training programme that we want.
02:40
Once we're happy with everything, we can approve and send the training programme to our client, saving us a whole bunch of time The third flow that I'm particularly excited about is AI-enabled onboarding.
02:49
This is a process that will typically take up a pretty pretty big chunk of my time and most of it was admin work.
02:55
So here when, I'm setting up a new client, I can simply add in the transcript from my initial consultation with the client or any notes that I took based on our initial conversation and I'm giving all of this context to Coach OS.
03:06
Then based on this, we usually want to build a training programme for the client to get started on.
03:11
So we can take all that context and we can make adjustments to one of our foundational training programmes that we have set up, or we can build a training programme from scratch.
03:20
Here I'm going to use one of our foundational training programmes.
03:23
Coach OS will use the context that we gave it and, adjust that training programme to suit this client.
03:28
You can see here that we extracted some insights from the transcript, such as the client's goals, any limitations that they have, preferences, experience and so on.
03:36
And then based on that, we made adjustments to one of our foundational training programmes to build the training programme for this client.
03:42
And again we can make further changes manually or we can request further changes using natural language.
03:47
Once we're happy, we approve and send to the client.
03:50
And this would have saved me an hour or so of manual admin type work.
03:54
Whereas now I know what the end output looks like, I can simply tell the AI and continue making changes until I have exactly what I need.
04:02
need.
04:02
These AI native flows enable me to save a whole bunch of time so then I can spend more time on getting new clients or providing value in a different way to all of my existing clients.
04:13
But again, I'm not spending all of that time on admin implementing changes in spreadsheets or apps.
04:18
And then looking at the client side, we're redefining how they interact with a coaching app by giving them a natural and frictionless way to track.
04:25
Firstly, you want to reduce cognitive load on the clients by giving them two main entry points.
04:29
They can either train if they're in the gym or they can track in a natural and unstructured way.
04:33
So you can see here they can either use a voice input where they can just pick up their phone and brain dump how their day is going or any specific trained performance metrics.
04:41
They can take a photo or add a screenshot of their training log for that day or they can text input if they prefer that.
04:47
We also interact with the apps that your clients already use, like Oura, MyFitnessPal, WhatsApp, where where your clients can, for example, message our WhatsApp integration with some unstructured input about how the day is going or that morning's weigh in.
05:00
We can pull structured data from Oura, MyFitnessPal and combine that with all of the unstructured data that your clients track and present patterns and insights to coaches.
05:09
As well as this, we'll have a, Siri integration so your clients can pick up their phone and just track on the go.
05:15
Or if they're in the gym, all they have to do is click train.
05:18
They'll be brought to their training programme where they can track manually the good old fashioned way.
05:22
Or they can track using natural language and voice input and just talking through the weights they lifted, reps RPE and so on.
05:30
Thank you for watching and I look forward to hearing from you.