Ole Shved is the co-founder of StackBob, a modern identity and access management solution that uses AI agents to integrate with any web application without requiring API or SSO protocols. We discuss how StackBob is disrupting the IAM industry by eliminating the "SSO tax" that forces companies to upgrade to premium plans just to enable single sign-on features.
The conversation explores how StackBob's browser-based agent can discover tools across an organization, synchronize user accounts, and automate provisioning and deprovisioning processes. Ole demonstrates the platform's interface and explains how it provides comprehensive visibility into tool usage while maintaining detailed logs for compliance purposes.
We also dive into StackBob's vision to become an "AI MSP" that serves as an automated IT administrator, the future direction of security software in the AI era, and Ole's insights on team building for founders. Whether you're struggling with complex access management or looking to enhance your security posture, this discussion offers practical insights on modernizing identity management.
TimeStamps to Jump in:
(00:01:11) Introduction to StackBob
(00:02:47) Experience with Okta and its limitations
(00:04:58) Why StackBob was created - problems with traditional IAM
(00:09:24) Why innovation has been slow in the IAM space
(00:10:57) The opportunity in simplifying enterprise technology
(00:15:58) StackBob demo - discovering tools across an organization
(00:19:51) How the AI agent works - demonstration
(00:29:06) How the AI agent handles provisioning
(00:32:29) AI as a single point of failure - how StackBob addresses that risk
(00:34:46) StackBob's vision beyond IT - serving security and finance teams
(00:39:59) Expanding into security concept and third-party risk management
(00:41:17) The future of AI in the workspace - AI is "under-hyped"
(00:49:12) Three possible futures with AI - democratization of software
(00:52:57) The era of the knowledge worker
(00:57:50) Target markets - small businesses to enterprise
(00:58:53) StackBob's current customer base and approach to growth
(01:07:28) What roles might be on the "AI Chopping Block"
(01:10:28) What's lost by automating IAM - "Nothing worth keeping"]
(01:11:54) Founder lessons and advice for business leaders
Referenced:
Check out StackBob
Try StackBob’s new Agent-based IAM integration feature
Find Ole Shved on LinkedIn
Transcripts
Find transcripts of all prior episodes here.
Ole [00:00:00]:
So there is new possibilities. Some jobs will go extinct, but new will be created, new, new demands. But it's difficult to say whether the bill will be same level of job available, more jobs than today or less jobs like we don't know.
Derek Skinell [00:00:22]:
Welcome everybody to the AI chopping block with home base. We are here with Ole, the co founder of StackBob. Ole, how are you doing? Where does this podcast find you?
Ole [00:00:33]:
Sure, we also have Newton here. Glad to be here. And my friend also I was expecting David to bring his dog because I saw on one of the previous podcasts you had this nice thumbnail. But all good, all good. Really happy to be here.
David [00:00:54]:
Yeah, it's in the other room, you.
Ole [00:00:56]:
Can still bring him.
David [00:01:00]:
Welcome man.
Derek Skinell [00:01:01]:
So Ole, why don't you, for the listeners, why don't you give us a quick rundown of StackBob, kind of how you thought of the idea, what value provides and where it fits in the space.
Ole [00:01:11]:
Absolutely. So StackBob is a modern identity and access management solution that is changing the way identity and access management industry works. And we are introducing agent based IAM to the market which means that you can now easily set up identity access management for your organization without depending on API and SSO as a protocols, but using instead our agent that is able to integrate with any web application you might be using, even including your custom build. So if you build something in house, no problem, you can teach the agent how to manage that tools and at the end you will have 100% of your tools covered centralized in a single directory where you will see all of your users, those being your employees, partners, so basically your workforce, but also non human identities. So any APIs or other agents that you're using in a single system and then you can control which level of access they have to your tools to your data, you can automate provisioning, so onboarding of users, monitoring them of usage and also onboarding of them. So that's what we do, we basically changing how the IAM industry works and our main competitors on the market are Okta, Pink, Identity Entrance and other SAML based solutions.
David [00:02:47]:
Amazing man. So the reason you guys caught my eye is because a couple years ago or a few years ago I owned the company. It's a staffing agency, staffing, recruiting and we built some product behind it that allows us to automate a lot of the things, but we used Okta or we were hoping to use Okta to do our provisioning. And so long story short, I've talked to the reps there, they told me I have a perfect use Case they even brought in a sales engineer. They said it's going to work well with all the tools that I have. After I signed the contract, I hired an Okta consultant. We've tried everything, it's not working, we can't implement it. I go back to them and I'm like, dude, it's not doing what you told me it's going to do. I can't use it, so I'm not gonna pay for this or I don't want to pay for this or can we do something here like reduce the bill? And they were like, yeah, yeah, we definitely want to work with you. They even got like a VP of sales on the call and a few days later they sent me an email and they're like, hey, unfortunately you signed the contract so you're gonna have to pay for it even if you don't use it. It was literally like few weeks later, I've already paid the consultant. They're like, no, I've tried this. Not working. So long story short, I've been ever, I've, I've been pissed ever since at this company and they keep emailing me every couple of months, like I think every time like the quarter approaches and then they need more sales and they're like, hey, you know, like, you know, are you looking to upgrade or maybe add some new services? I always like respond with like a really angry email and the emails keep coming. So long story short, I. Your product looks really interesting. I never implemented Okta, but you know, like, maybe we could use it or maybe you can tell me, is this even an Okta competitor or is this something to be used with Okta? What's your big vision? How do you see the world changing with AI agents and why even build this when there's Okta out there?
Ole [00:04:58]:
Yeah, sad to hear this story. Absolutely. Okta seems to be powerful, but at the end when you really get into details and you try to implement some specific scenarios, you often will bump into the limitations of old protocols, which is being SAML based sso. And yeah, so this is not a unique case of yours, but said that you had this experience and this is one of the reasons we are doing this. So to give you more power and more automation capabilities, more granularity in your access management, in your provisioning. But the main reason we're doing this is because, so to say very good phrase, I got today actually from the previous call, the poverty line in security is quite high. So if you want to set up the iam, you need to be really rich and you really need to invest a lot of money and effort into setting everything up. And the one reason why it's like that is because there is kind of cartel situation on the market which is also called SSO tax. So if you want to set up single sign on sso, you need to first upgrade your every single tool you're using in your stack to their premium plans. Then they will allow you to use SSO as a feature and it's a cartel situation because everyone is doing that on the market and it's a standard. So Basically imagine you're 100 employee company and you want to have centralized IAM solution. Then you start looking for something, you probably will end up with Okta because they are most popular solution on the market. And then only then you will realize, oh wow, in order to use Okta, I first need to upgrade every single tool I have to premium plan, which means I will start paying like two, three, sometimes four times more money for subscription than I did before. And even if the basic plan is finally totally fine for me, I can't use it with Okta because the SSO is hidden in the premium. So this situation on the market is really unfair and we believe it shouldn't be like that. But this is the one problem is the cost, so the assets of tax so to say. And the second problem is that the Okta and similar solutions are really complicated. And as David, as you said, you really need to hire a dedicated person who is specialized in Okta setup and spend a few months, typically like three and more months on setting everything up. But we even talked to some series A startups who spent a year migrating to Okta and at the end they still were not happy and it was very painful process. So the second problem is the complexity and time to value. So you really spend a lot of time before you get value out of that. If you, even if you do, maybe you don't like that in your case might be that it will not work out. And the third problem is actually the coverage. So not every tool has SSO as a feature and even those tools that have SSO as a feature, they still might be very limited in what you can do through that protocol because it's quite legacy protocol and not everything is possible through that protocol. So you still will be doing some stuff manually even you spend all that money on upgrading every single tool, you'll spend all that months migrating to and setting up something like Okta and then still at the end you will have spreadsheet with some tools and you're managing manually. So this is madness. And that's partially because the last innovation in the area was 15 years ago when everything was moving to the cloud. And then Okta appear as the first cloud based IAM solution. But since then there was nothing happening and we saw this opportunity on the market specifically with AI providing this capabilities today where you can teach the agent to understand different apps and to automate different actions in those apps. So we see the AI as the next step and next revolution in IAM as it also do in other industries obviously.
Derek Skinell [00:09:24]:
Ali, I'm curious, why hasn't anyone else innovated here before? Or maybe there are other innovators. I. I will admit I'm not an IAM expert or enthusiast, but yeah, how, how come? So Okta seems like they're sitting fat. Sitting fat and pretty and breaking in the dough for not innovating what's right.
Ole [00:09:42]:
Yeah, great question. So it's typical situation when the larger companies and big enterprises, they are very slow to innovate and that's always like that. They are maybe they're trying to do something in house, we don't know. They're not public about that. But as it typically goes in history like it takes some time and very, very few cases of successful innovation inside huge corporations. So typically they will be looking for startups who do the innovation on the market and then just do M and A because the startups have this drive and this motivation to change something, to build something completely new, to take those risks. And then those startups that become successful, they just being acquired by huge corporations and that's how it works.
Derek Skinell [00:10:32]:
Okay, and are there other incumbents in or new entrants in this space that might be competitors to stack?
Ole [00:10:40]:
Bob, there are a few companies that are trying to innovate in iim, but they're very early stage. So there is no like big competitor or anyone who is far ahead. So it's very early stage still.
Derek Skinell [00:10:57]:
Okay. I remember here listening to another business podcast. I don't remember where what it was, but they were talking about how one of the big opportunities for companies moving forward, especially in the age of AI is not going to be your innovative technology but your ability to cut through red tape and bureaucracy and bullshit. And that's, that's a huge opportunity for folks. And it strikes me that this might be one of those things right where SSO and IAM is like it's not necessarily bureaucratic red tape, although it could be, especially with larger enterprises because we're talking about the security of your access to all your technology stack in your database, but also trying to figure out how the hell to integrate to all these platforms and manage a single sign on for all this. So I'm curious if again, consider me green to the IAM space, but how close am I to the mark here? Is that kind of where you see your opportunity and your, your wedge?
Ole [00:11:52]:
Yeah, it's very, very, very nicely put. And also I wanted to add that to your previous question. So the security and iam, it's, it's a difficult space. Right. So you need to have a good understanding of technology and of the market. So I have, I'm lucky to have a great co founder who is my master of science in cybersecurity and extremely good engineer with whom I was working at the previous startup where we're building enterprise grade software used by Nike, Coca Cola, Uber and other big brands. And we had this insight about the insufficiency and like the problems with IAM working there and we were brave enough to step in and say hey, are we going to change this? So actually the entry to this, like to this nation, to this market is much more higher. So the threshold rather than building like sales automation tool or productivity tools. Yeah.
Derek Skinell [00:13:07]:
And so you said your previous company is that Smartly or Xenos?
Ole [00:13:11]:
It's Smartly. Where, where we were building the dynamic creative optimization tools and actually it was part of Ad Libs. So this is a startup that was acquired by Smartly, which we were building from early stage from basically from MVP to acquisition by Smartly. And after acquisition we started working on Stack. Bob.
Derek Skinell [00:13:36]:
Gotcha. Now you said your co founder is data scientist. Are you, is your background in like IT and SSO and IAM or what's. How did, like how did you get into this? Was this always something you were working on or did you just run into an issue one day and you're like I gotta, I gotta fix this, it's driving me nuts.
Ole [00:13:55]:
So before Stackbox I was working in five different startups in technology. All of them are B2B. So my background is a B2B software building and not specifically cybersecurity before Stackbox. But I was well aware of the problem with how big businesses are managing accesses. And it was my personal problem at the previous projects where we were growing a team very quickly from like few engineers to over 30 engineers. And this access stuff was my personal problem as well. So that's how I having a really great co founder with background in cyber security and high load and data science engineering, how we end up building. So I'm more business person but with a good understanding of the problem while my co founder is great technology person.
Derek Skinell [00:14:57]:
So you said that the barriers to entry in this industry are really high but it sounds like to me you have a really compelling value proposition given how you put it, the high poverty line within it because you got to upgrade all your solutions to the enterprise plan to be able to access SSO costs a lot. So if the barriers to entry are high, I guess where what do you think your secret sauce is that's going to allow you to deliver on this value prop? And I have a feeling it has something to do with AI agents because you've mentioned a couple times and that's what we're here to kind of evaluate on the chopping block is what type what's not, whose job is on the chopping block. So maybe you could tell us a little bit more about that secret sauce, how you're working AI into your product. Because for life me I can't, I mean I like to think I'm a pretty intelligent guy, but I can't figure out how AI is going to help manage sso. Is it like going in and typing in my password for me and logging me in? Like how exactly does this work?
Benjamin [00:15:58]:
Also too if you. I'll add on to that. Like obviously there's been stuff that's being released every month. So how did you guys go about selecting your tech stack? You don't have to give away your, you know, secret ip but obviously if you're using AI agents, you know, how did you go about selecting that and making sure that you landed on the right.
Ole [00:16:17]:
You mean for new stack for development or what? Exactly.
Benjamin [00:16:25]:
So, so you have AI agents working with Stack, Bob. Right? Yeah. So I'm talking more on the AI agent side. How did you go about selecting that tech stack and making sure you're, you're, you're selecting a winning horse there that you're not going to run into problems.
Ole [00:16:38]:
Great questions. So our secret sauce is exactly the using agent in order to integrate with the tools and to automate the actions in the tools. Rather than using traditional API or SSO integrations which is browser based agent that is able to first integrate with any tool and collect all the relevant information about your user accounts, their level of access, their activity, their privileges and also automate create rules for role based access controls where you can create basically for new employees or new people joining your team. Automate provisioning of those different accesses and licenses and also then deprovision them when the person is leaving a team. So this is what differentiates us from existing solutions and for the tech stack question. So we build the core agent which is able to execute the assignments that they have, mainly like provisioning the provisioning or collecting the data. And using AI here helps us for the agent to learn new tools. So obviously there are like thousands and hundred, probably like hundred thousands tools on the market right now for different teams, different needs, a lot of competitors, tools doing similar stuff, but there are really a lot of them. And our offering is that we can support any tool, it doesn't matter what stack you have. We can integrate. Our agent can learn and be trained to integrate with any tool that you have and automate the provisioning, the provisioning of user accounts in those tools. This is the AI part where the agent is learning the tools and being able to automate actions in them. Even if you have some custom tools that you've created in house so that you can train the agent to integrate with your custom tools as well. So this is the main part. So basically it's not kind of traditional generative AI part where it generates something for you as probably most tools on the market, it generates the integrations which are the things that are working under the hood for you. So you are not using the product of AI, but the agent is using this generated like workflows to deliver value for you.
David [00:19:22]:
Is there a demo that we can see? Because we've seen some AI agents. For example, the last call that we did, the LinkedIn AI agent will go and basically select a list of people to target and then we'll start and comment on their posts and stuff like that. Is that kind of like how the product works? It like kind of controls your screen and or not controls but like kind of like, I don't know, takes over your computer and starts inputting stuff or how does it work? Like is it possible to see?
Ole [00:19:51]:
Sure, sure, definitely. So I need to share my screen which is here, here we have this is demo environment but it will be like believe enough for us to go through the main part and we probably going to go through typical flow that the user goes when start using our product. So the first part is we help you. So the StackBob, the product helps you to discover all the tools that are being used by your team. And this is helpful for those companies that don't have any centralized place where so those like spreadsheets or so to say where they keep track of everything. And there are a couple of channels how the tools are discovered. So the first through the agent. If you or the people in your team install the agent, then the agent will collect the user's data from the browser itself and it will discover the different workspaces. So you might have like multiple workspaces of the same tool. Not a problem. Like it's going to discover that. And we have a more traditional channels for discovery which would be discovering payment transactions. So if you connect your payment provider, even there are tools that you haven't been used for a while, then the system will still discover the tools that you're paying for but you forgot like nobody is using. And then there are traditional integrations for Google and Microsoft where even if the team member doesn't have the agent installed on their machine, then still the platform will discover tools being used by your team members through Google, like where they use Google login or Microsoft login. And there are some other channels of discovery of tools we are working right now, so it'll be even more soon. But this is like the current version which is quite sufficient and it covers like 85% of the tools that your team is using.
Benjamin [00:22:11]:
And this is your tech stack that you have.
Ole [00:22:13]:
Right?
Benjamin [00:22:15]:
Right.
Ole [00:22:16]:
So this is a demo account, but some tools for you as well.
Benjamin [00:22:21]:
Demo account. Okay. I was going to say throw David's in there so we can see all of his subscriptions, everything he signed up to his tech stack to run his company. I was going to say throw David's. Throw David's company in there so we can see everything David's running to run his company. Absolutely.
Ole [00:22:44]:
Yeah.
David [00:22:46]:
I'm a sole proprietor. You don't want to see what's under my hood.
Benjamin [00:22:52]:
But this is all, this is all internal. You have to log into the solution.
Ole [00:22:55]:
Right.
Benjamin [00:22:56]:
You have to log in and it kind of goes through this discovery. How does this kind of like reduce the scope? Like let's say me at an enterprise company, we have, you know, 122,000 employees. We have a massive procurement tech stack right. Where we can just buy software like crazy. Is this going to go company wide across the entire domain? And how do you make sure that we've captured all of that at a company like that with that amount of complexity?
Ole [00:23:23]:
So first thing is the power user of the tool is the IT department and also security department. And there are two ways how you can adopt Stack Bop in your organization. So the first you can just use it for IT and security teams and don't distribute it across rest of the team. And it will help you to manage those tools that you centrally manage. It will Be less of shadow of it, but still it will help you discover some shadow of it through Google and Microsoft integrations. And the second option is where you want to distribute StackBob across your entire team and make it your like security and identity like single sign on provider, which is currently in a beta, but we're going to roll out in probably closer to the next end of the next quarter the SSO like solution. So right now this is more identity index management for admins which allows you to automate provisioning and deprovisioning of the tools and keep control of everything. But for end users we offer right now just enterprise great password manager, which is more traditional way of keeping track of passwords. But the next big step and next big innovation that we're going to release will be SSO like so single sign on like solution which will use very special approach of creating sessions in Stackbox and then sharing them with your end users without them being able to use passwords. And we're still thinking how to package basically that part, but that will come soon and we're probably going to talk about that a bit later. So right now the main value of the product is that it allows you to integrate any tool. And we can take probably Atlassian as an example, an SIT or security team. You can synchronize the tool and get list of every single user account you have in every tool that you use. Then for every user account you'll have list of licenses or products that were provided to this user. You'll have the permissions that were granted for this user, you'll have the cost and the user's data. And from here you can also do the removal. So you can deactivate the entire accounts, you can remove specific licenses, you can do like bulk actions and remove bunch of users from a single tool or from multiple tools. So basically do your provisioning and deprovisioning and we have the view. Like you can have the view by tools, you also have the view by users where you have all of your human and non human identities that are currently managed, that are being provisioned, that are being deprovisioned and also your former employees or partners that's already been deprovisioned. And then for every single user, this is a demo account, but let me find someone with some tools. Maybe this one has few. So in real life you'll have like tens of tools for every single person and then you can also control access to the tools for every single tool or you can do the scheduled deprovisioning. So you can say how this user is leaving us at the end of the month and we can deprovision him entirely. Then you have the view where the system recommends you all of the inactive or dormant accounts that you have across all of the tools. Here we have example, we have this user having been using the tool for 12 months. And here we have like Figma, Atlassian, HubSpot and other tools. So you can also review those recommendations and then do the bulk removal for account that you believe should be removed or you can create some rules for that as well. And then you have all of the logs across the organization or you have logs across specific. So every single tool where you can see where for example for this workspace when the access was removed for a specific user account or when it was provisioned so granted or requested. So everything, every single action is being recorded. And this is important for compliance audits. So yeah, I know this probably was too much information. So please ask any questions you might have based on this.
David [00:29:06]:
How does the AI agent like for example, you know, like I just hired a new engineer, how does the process like how does this work? Does the AI agent control someone's computer and then walk over to Google workspace and create an account or workflow? How does AI agent play a role in this?
Ole [00:29:34]:
So the provisioning part is if you don't have any automation solution, then your admin would do it manually, right? So you don't need to the end user, the person for whom you want to create the account or remove the account to do anything. So in alternative scenario it will be admin who will be doing this manually. And we offer a solution that automates this. So agent will be doing this instead of admin. So you don't need to have end user being involved. And right now this is agent that helps admin to automate these actions and the agent works. So it's browser based. And for example we can do the first very simple action which is synchronizing the tool. So when I'm syncing the tool I just trigger this and then the agent I have in my browser. So this is a StackBob agent I've installed here. It triggers the process and it opens this synchronization window. So now it's in this specific demo environment in using my credential and running this on my machine. But it's possible to run this on the server. So it can be done without my involvement automatically by triggering this based on like timeline or other triggers. But right now for demonstration reasons, this is run on my machine and then so the agent is going to this specific Atlassian workspace so Demo Ink workspace. So I have the URL which is for this workspace and then the agent is collecting the data for user accounts that are in the system. And now we see it was just saying less than a minute ago and now this all information is up to date so we can see okay this Robert and Dylan were using atlassian in last 24 hours. Eric was also using it recently but Julia have been using it for five months and Big Bob never use it and from here in the same way so that we can trigger like removal. So if we if we're going to remove Julia then the same way the agent will trigger the process will open the window and just go find the specific user account and remove and deactivate the user from the system or provision the user. So basically based on the on the task.
Derek Skinell [00:32:06]:
Ole, I'm curious a lot of times when we talk to other founders with AI solutions there's still a lot of human in the loop and so I'm curious how do you prevent AI from becoming a single point of failure within StackBob, in terms of managing. Yeah, managing like identity access management.
Ole [00:32:29]:
Ideally like in the long run we believe this would be your AI admin so to say. So if we're going to talk about the team of your podcast, so which jobs are under risk? We believe this job of like provisioning, deprovisioning can be fully automated and managed by AI. Obviously you would still have some chief security officer or chief Information officer overlooking this and defining the rules either by communicating with tool or just by using a regular interface, whatever they prefer. Because I also see a lot of companies trying to use chat as a main interface for interacting with the tool. But honestly quite often it's much more hustle like writing and typing and communicating rather than having a proper interface where we can clearly see what's going on and just like turn on and turn off some things which would be much easier. So it depends on the scenario how exactly you're going to interact with tool, but in the long run this would be like your AI IT administrator that will do all the provisioning and the provisioning and we see right now in the market there are a lot of MSPs who offer managed services for your IT tasks and we believe the StackBob can be your AIMSP in the future where you don't need to hire additional third party consultants that would do this manually for you. But you, you're going to have your AI based IT admin so to say in House that will be responsible for that.
Derek Skinell [00:34:21]:
And so you've, I guess where, where does this stop for you? Then you, you mentioned like you see Stackbot becoming kind of your, your AI admin, at least for provisioning and access management. But are you looking to take over more of the IT department and support support department there? Like where, where does stackbob stop taking over the world? Or is it all of it? What's kind of your long term vision?
Ole [00:34:46]:
Actually even right now it's not only for it. So IT and security are primary users. Primary users where we for IT we automate the workflows. So the manual tasks they do for security teams, we provide the full view of all of the accesses and permissions and we help them to streamline the compliance process. Right, because if you want to go through security compliance like SoC2 or ISO, then you need to have IAM system in place. You need to be controlling your accesses centrally. This is a requirement. You can potentially do that in a spreadsheet very regularly in manual way, but that's not going to work in the long run. So this is our main value where we target IT and security teams. But because our first MVP was actually around the license management. So we also have right now the module which helps you to cut your costs by identifying, removing redundant licenses. So we also interesting we provide value for CFOs or COOs, the people that are interested in optimizing the spend at the company. So this is our second champions or second ICP inside the company. And we also at the end we help every single team member with like not hustling with passwords, not having those hard times with shared accounts, specifically where you have maybe some social media or other like very expensive databases or other services where you single account per company or single account per team or per few people. And StackBob helps you to manage the passwords. Right now it's more similar to traditional password managers like one password. But in the future it will be more complicated and it will be more secure by not allowing the end user even to interact with password in any way or being able to copy it. Because right now if you use 1Password, for example, you can copy the password and then still share it with other person through WhatsApp or whatsoever. But we are building the solution that will keep the passwords out of the user and just give the session to the user. And at the end it will be more like sso, like experience without actually using the old protocol. Right now this is, we offer you SAML SSO capabilities for Provisioning and deprovisioning which is a huge part of why assets are being used. Because you can also use like Google Workspace or Microsoft for logins asset so to say single sign on. But it's not the real like saml how to say so the SAML based assets. So it also allows you to create accounts and remove accounts which is important for security reasons and for keeping track of all that actions. While if you use just Google login or Microsoft login as so to say like social login which is for end user, it feels like single sign on because I've logged in with my Google and it lets me in into every single tool. But for my IT admin and for my security guys it doesn't help them because they cannot do anything through Google, they cannot create accounts, they cannot control accounts. And this is a problem we are solving right now in IAM where we automate provisioning and deprovisioning and control of every single account and every single tool. But in future we also will have this for end user, even for tools that don't have SSO at all and that currently have passwords.
Benjamin [00:38:41]:
One of the things I'm interested to know about, you know, beings that you're in the security world is you know, okay, so you're solving for this SSO stuff but have you thought about like the next step, the logical next step in security where it's like oftentimes a company will say hey we're really interested in this solution, this software and we want to connect it into CRM or you know, our Martech stack. So we want to bring it in and start connecting it and then you get this whole world of security concept third party risk management assessment forms. You, you get the security concept and maybe hacking. So are you looking to take it a step further with security along those lines where because you said that your solution kind of learns the tools that one has. So like are you going to get into this world with AI where you're learning like okay, this is how you're going to connect and here's all the forms you need to be compliant to connect. This is some of the data processes, things that needs to happen and then obviously you get into provisioning with that, right? Is there any world where that happens with your solution? Because if, if it is, I'll sign up right away. Right? Because you know those security concepts are.
Ole [00:39:59]:
Monsters and sign you up.
Derek Skinell [00:40:02]:
So your salesman.
Ole [00:40:04]:
That's my job. That's my job.
Derek Skinell [00:40:06]:
Yeah, but we'll make promises and fulfill them later. It's all about getting, getting them on board. If they sign the check, then we'll, we'll figure it out.
Benjamin [00:40:13]:
No, because that process, I know in some enterprise companies not, not saying any names but you know, just to get a security concept done and go through all your checkpoints to say, okay, we can actually connect this tool to that tool and get all the security requirements out of the way. And obviously access is a big part of that. It can take you eight months before you touch one configuration in the tool. So obviously that's a pain point that I've been thinking about and I've dealt with for the last five years.
Ole [00:40:44]:
I believe we need to make a separate discovery call with you on that.
Benjamin [00:40:49]:
Yeah, there you go. There you go for sure. No, that's awesome. Where do you see like AI going in the next couple years in the space? And, and do you really think that Truly, like it's gonna. Because they always say oh, AI is going to create these, you know, all these jobs. Do you see that happening with AI in this space? Is it going to truly create more jobs or is it just going to allow guys like you to get started up with less employees?
Ole [00:41:17]:
That's a good question. I think like overall AI is probably under hype. That's my opinion. Because under hype because like truly it brings so many new possibilities in what can be automated and what will be automated that it's probably difficult to digest for humanity. And obviously in attack space everyone is talking about AI and implementing AI but like in other industries people like probably talk to Gemini or ChatGPT sometime but they, I'm not sure that they are thinking like how much will be digitalized and automated in our day to day lives in the next like 10 years. And I believe it will be like really huge changes and like personal Jarvis's like in Ironman. I mean it's, it's a no brainer to have one. I mean I was dreaming of one like for a long time and now it looks like it's gonna be reality very soon. And also then the software, right? So the, we had this shortages of developers, extremely high salaries intact compared to other industries. And now like creating a software tool is becoming really, really simple with what AI brings to the table. And honestly the software becomes, I really like the comparison where the software becomes new content. So before a lot of people were able to create content because they know how to write. And now a lot of people will be able to create their own software for any need or problem they will have. And this is probably true for like personal problems, but it's more relevant for businesses. So if you have, if you work in a big company, you have tons of different needs and problems that you need to solve. So today actually we live in the era of the highest abundance of tools that we have on the market and the highest number of tools that are being used in companies. But I believe it'd be like much more because now every single individual or teams will be able to create their own tools for some specific needs they have. And it will be like just like explode the number of tools being used. And that's why we also creating StackBob to be ready for this changes that will gonna happen because with traditional technologies like with Octa, like you can do nothing with those like other agents or other custom tools that were built, but the agent we are building will be able to handle those as well. So I believe like there are really a lot of changes coming to the every industry right now. And probably it's difficult to imagine how the future will look like in 10 years. But to answer your question, like will there be like more jobs created by the AI? Like it could be similar to what happened with content creation, right? So we had previously like few big media outlets that were creating content like the TV channels and so on. And now we have micro influencers, YouTubers, TikTokers like so many people and having their smaller audiences creating the content. It could be similar thing with software so that people can create their own like flavors of tools and then have just like very small audiences. So there is new possibilities. Some jobs will go extinct, but new will be created, new demands. But it's difficult to say whether the bill will be same level of job available, more jobs than today or less jobs like we don't know.
Derek Skinell [00:45:39]:
So Ali, you touched on something that Ben and I have been talking about for a while and I remember bringing this up on a previous episode. But I kind of see three paths for the world in general with AI. The first one is AI takes all our jobs and all the power is consolidated in just a few select companies, OpenAI and Anthropic and Google. And we just live in this like dystopian cyberpunk tech oligarchy and they have all the power, they have all the AI. And on the other scenario it's this utopian world where AI takes all of our jobs for us, no one has to work anymore, we all get UBI or, or some payment, right? Or they fix scarcity and we can all just like sit around and paint and make art all day and we could do whatever you want. You can program for the fun of it. Both of these seem, well, one seems more likely to me than the other and I won't, I won't speak on which one. But there's a third option here which you've kind of touched on and I think it's the fragmentation of the market and the democratization of software, right? Because you can now vibe code, which I love, love the term, I know real developers like hate it, but whatever. Now you can vibe code your way to any, any application, right? And they're, I mean they're pretty usable. We even see people like you said, software is the new content. Every 20 year old kid who wants to be an entrepreneur on YouTube can just spin up a landing page with AI and then for their product they've built with AI and they're just ripping off one or two features from an existing legacy product and they're selling that for like 10 bucks a month. And so the way I see this going is, or the third option here is a world where AI democrat democratizes the building of software much like, like laptops and personal computing did for content creation, music and video and game development to a point where everyone, like you said, can build your own apps and you have your own small customer base. And I imagine a world where maybe, you know, for me, I'm in Denver, I have Derek's software house and I'm in the third floor of a building and in the fourth floor is my customer. It's a marketing agency and they're providing marketing for the bakery that's on the bottom floor. And I provide the AI tech stack for them. And the database provider is in the second floor right below us. And I run my AI on their, you know, database and I'm just, just like the coffee shop and the bakery serves the businesses that are on that block. I serve my software to those people, right? And the cost for me are low. It's made me and one other person, me and Ben. And of course we use StackBob to manage the access and provisioning for all these apps. But you know, I see a world in which we all become blacksmiths and cobblers again where we specialize in our craft and we're just building software for the specific people either geographically or we organize, self organize some other way. But there can be a hundred thousand ERP software, there can be a hundred thousand sales automation CRMs, whatever you want because everyone's going to have a slightly different way that they want it done and there's going to be someone else out there that's like, yeah, I can like build an app with AI and manage it for you. It's about as hard as baking this bread or, you know, cleaning your car at the end of the day. The effort now and the cost to do this is so low that I'm doing it because I enjoy it, not because SaaS or technology is this outsized return. Right. I think it really democratizes the playing field and devalues everything to such a point that it's all worth the same. So this is, this is kind of my thought experiment. Lately I've been throwing this out there in conversation. All my wife's friends hate me now because I won't shut up about this. But I'm just curious to get your thoughts on kind of these three outcomes.
Ole [00:49:12]:
Yeah, the last one resonates the most with me and it's like local communities or maybe it will be not local, it could be just by interest. But I believe this community building and brand building and common values, common approaches will play a big role because the software will become the commodity and anyone can build it and then it will be really difficult to differentiate and then creating a community and whether it's being like local or some other interest based will be a big player in the near future in order to win customers and keep customers and attract customers. So I believe your third options is quite viable and at least for the next few years, and then maybe in 20 years we'll end up with something else.
Derek Skinell [00:50:19]:
But the pendulum will always swing back and forth. Right? But yeah, I'm kind of encouraging this model. Right. And I think, I mean, honestly, like, I think all the big legacy dinosaurs, as big as they are and as old and slow they are to innovate, I think if they're not going to innovate, they deserve to go to the wayside. And some of these companies have been around. I mean, Google's SAP is 40 years old. Google is 30 years old, roughly 25. I mean, in technology speak, that's a long time. My Mac, my MacBook Air from 2020 is now four years old. That's an M1 Mac. It's a dinosaur now. So, yeah, I mean, I, I think it's time for fresh faces. And, and it also, you can call me a bleeding heart liberal, but it also redistributes the wealth too. Right? Like does, does Google need necessarily need all that power? And I think it's great to put that back in the hands of, of the many. There's gonna be a lot of crap that's created, don't get me wrong, like, and we're going to have to sift through a lot of bad stuff, but there's also going to be some really ingenious things that are created. And I think about, I'm a musician and producer and I really like to discover new music but it becomes harder and harder because more people can make music, it's easier to do so it's really hard to find good music. But at the same time more good music is being made if you know where to look for it and how to find it than ever before. So you know, if you take the good with the bad, there, there's a lot of opportunity for everybody. So if there is anybody out there that's like, oh, there's too many CRMs, I had an idea but like I don't want to create another CRM. I think you should create it. And I've really come around to this idea. There's no reason that you can't create StackBob V2. Sorry Ole, but someone out there might say, well I think I have a different idea or maybe a different UI or different workflow and someone else is going to prefer that to StackBob. And that's okay. There's room for everybody.
Ole [00:52:02]:
The stack up is much more difficult to create than the CRM but you can try.
Derek Skinell [00:52:06]:
Yeah, that's true, that's true. You have a better moat.
