Dinis Cruz  0:00  
So, so on the ABP policies, actually, there's another very interesting angle that we can add, which we touched on it, I think. But we actually, I think that needs to be another another primitive that could also be solved, which is use case driven, and I think what's interesting about this use case driven is that we can then start combining several policies. But this is where we then combine, for example, very specific. This is okay because remember that we are providing specific examples that then the customer can pay for their version, right? So, take for example voice debrief. So this is a case study that we can say where we it's basically we have multiple scenarios. In fact, we can create multiple policies for voice debrief that focus on specific workflows because we already have very specific nice workflows that we do there, which is also a great way to advertise what it does. So, for example, we have the web flow, which is when you go to the web, you you then use. Actually, this is interesting because what I'm about to describe is that you have a policy when you have open router where we don't know where the data is. So it's a good example of governance. So if you look at this workflow, is that if you look at voice debrief, you go to the website, you upload your audio. We we then say transcribe, get infographic, and that now is sent to open router multiple models, which then operates the data on it, and then sends the stuff back. So, for example, the the grant is that the data can up anywhere. The mandate would be: I don't want anybody else apart. I don't want anybody else apart from this transaction to see the data. But we don't know where the data is. We don't know where the data is stored, especially especially this case because we're using a promota. We have no governance at least this stage, so that's one use case. The other use case is when we use, for example, WhatsApp workflow with NA10, which is the one that created there, which is the one that has all those examples from the thing we created. So this is where it becomes very, very focused and specific. But I think actually the way to do this, which is more interesting, is let's create the prompt for the agent that I have that knows how voice debrief works and knows how the NA 10 work workflow works to actually be the one that creates the policy, the grant, and the mandate that can then send to us and we can package it up. And actually, this actually opens up a very interesting workflow, which is we can have a workflow where we give the user a prompt for them to calculate what's happening in that environment, and then we'll take that prompt and we'll create the policy for you, and yeah, and that, and ultimately, actually, that's what that needle service is. So, if you look at the the work we add into the store, we're adding four price points. We're adding a price point, price point which is a zip file with all the policies. We are in a price point that you get a vault out of it, which is like the first is five quid, the second is 50 quid. We're having a price point, which is, you know, we create a custom policy for them. And actually, this is the workflow. The workflow is we give them a prompt, they run the prompt in the environment. They send us the result, and then and then we send them back the prompt, right? And then the fourth one is an actual interview, two sessions, live interview. So yeah, this is the first time we're operating this, right? So what we're doing here is we're showing in practice what does that 500 quid workflow gives you, and yeah, and we need a page, and we need to map this app, and we need to map the server. So this, so let's first capture these workflows, and the store skit. AI agent is working on this, so then we can synchronise with them once they finish this.

Transcribed by https://otter.ai