Enterprise AI tools need central control over access and deployment

Transcript

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

1. 10:45

   security you mentioned some stuff around governance yeah what makes this difficult yeah it's a challenge you know like i look i look at the feet of cios or the i put myself on their shoes and i think about okay i am a cio and i know that my company needs to use ais for sure right now maybe i have copilot and it works it's very good because you know it comes in microsoft i have chat bt or cloud enterprise good but okay personal ai tooling or personal ai agents can get you very far but when you look at more organizational applications of AI and you look at more process automation or transformation you look at systems that are not owned by one individual but even sometimes I was one team really like they operate cross-functionally and when you're building this you're truly building solutions or applications more than like personal interaction tools which

2. 11:32

   is like the chatbots so just applications can have many other forms can be can be like automations around the background can be systems that operating that operate through a form interface what not and people will probably call an application as their first intuitions how to solve this if they're technical enough if they are like into by coding a lot they can by code something and then the problem they face is that okay i have the hr team that built some random tool here i have the Legal team, I built some random tool here.

3. 12:01

   I have people here that, you know, collaborated between HR and support to build some random tool. And like, wait, wait, where is this hosted? Who can access this? Who can log into this? Wait, you know, who's giving them the credentials? Yeah, exactly. You have permissions everywhere. Yeah, no, but like, really, it's a problem. And then they get hacked, and then it's like, oh, I got screwed. And what we did is that, you know, what we've seen is that really, like, one of the concerns is if my organization is going to truly build that said a suite of agentic solutions and like to you know solve their problems and to make things easier for them i want to have a front seat at managing how is that happening i want to govern who can create something who can deploy something who can manage who can access whatever was created

4. 12:43

   and be able to upscale it downscale it or shut it off whenever i want to and moreover i want to be able to like open access as as as they have done with any other software and i think at the area of personal software came way too fast for enterprises and what we did was when we built a product that was first for that that you got the first hand so that you know if you're a big hedge fund that really cares about security and you don't want people to be building like random apps for uh you know trade monitoring okay i'll have a platform here where like you can build your app still no code it's very agent first very audible with analytics everything here and it will do so in a way that you can like use it in a collaborative fashion you can like you know operate across teams you can like easily inspect it because we're

5. 13:27

   designing a way that is easily no code so you feel like you're always kind of like designing a process together even though you're chaining these five skills I care about like using a computer interacting with some tools parsing a document and and at that point you have like a truly governed secure way of handling AI the other question people have a lot is like okay I'm bringing an AI agent here you know behind this a language model and this