You may watch the interview here: https://sedai.wistia.com/medias/j8w7ofs59s

Suresh Mathew (Sedai CEO): What were your top takeaways from autocon/23?

Rachit Lohani (Paylocity, CPO & CTO): First of all, congratulations on hosting all of us. This was a wonderful conference. It was amazing to see people coming from such diverse backgrounds, such different verticals, and all being tied together by Sedai, the platform. I had no idea that there was so much automation and autonomous talks happening in all these verticals. And it was amazing to see you have such large customers, that at this scale, you're able to support companies with hundreds of thousands of transactions a month, right? It's incredible.

So, back to the question, key takeaways. It was wonderful to see that there is so much excitement and energy around this concept of driving autonomous systems. But at the same time, I could sense a palpable consternation around how will I get there? What is the path to get there? And I think you guys did a really good job of tying the pieces together that will help, especially companies like me, pave a path from where we are today to what do we need to do to get to this autonomous system state.

Levels of Autonomy

Suresh Mathew (Sedai CEO): You talked about five levels of autonomy. So where are we, and where do you see us in the next three years?

Rachit Lohani (Paylocity, CPO & CTO): To summarize the five levels, it starts from L0 to L5. So we usually ignore L0 because that means everything is handcrafted and all manual. L1 is when you have infrastructure as code. L2 is when you have central automation systems that have some semblance or knowledge of the ecosystem, and then you can drive automation through that.

L3 is around observability, getting signals back. And essentially, conditional automation, which is scale up, scale down, and all those things. L4, L5 is around-- L4 is around mesh technologies. You can divert all your traffic proactively before you see a problem. L5 is autonomous, which Sedai promotes, right?

Most of the companies today, a good portion have graduated to cloud, but less lift and shift and other strategies. And they are slowly embracing cloud-native, cloud-first mindset. So we are between L2 and L3, marching towards L4, eyeing towards L4. The good part about all of this, Sedai as a tool, as an application, is not tied to a certain level. It helps you with expediting your journey towards this autonomous system.

Embracing Sedai for Autonomy

Suresh Mathew (Sedai CEO) Got it. Thanks for the first part. Now, you mentioned Paylocity is at a certain level. How are you embracing Sedai to move forward?

Rachit Lohani (Paylocity, CPO & CTO): Exactly. And this was very exciting for me. Currently, we at Paylocity, we must be-- so we are in a data center and moving to cloud and embracing more lambdas, embracing more cloud-native, cloud-first technologies, and the architecture.

But guess what? That is not an impediment. I can actually embrace Sedai and get to L5 in a very compressed timeline. That means the people who work on these systems would be free to work on better tasks that will help contribute towards better business. That means we get to put more towards R&D and not just towards running the business.

This was exciting for me.

Strategy for Autonomous at Paylocity

Suresh Mathew (Sedai CEO): What’s your strategy for achieving autonomy at Paylocity?

Rachit Lohani (Paylocity, CPO & CTO): Yeah, that’s a really good question. So when we started the journey at Paylocity, it was mostly around-- so we were in a data center, slowly moving our workloads to cloud. And now we are at a point where we have a good number of systems running there. Some of them are monolithic. Some of them are on lambdas, depending upon what the domain is.

We, at Paylocity, we move a lot of money. So there is a lot of compliance, a lot of rigor around what we do because we don’t want to send wrong paychecks to people, right? That won’t be acceptable anywhere.

And so given the composition today, we feel very comfortable operating the system manually. And we have the right signals that we get from our infrastructure that will help us make the right decision. And this is where we think tools like Sedai, which becomes a platform for how we operate, it would change the way we operate the system.

So rather than a person making decisions around, should I scale up, or should I scale down? Or I should restart a box, or how much JVM should I allocate, or how much memory should I allocate to this JVM. All of that goes away. All of that is not even-- not in a config, but in an automated system that has the memory to keep fine-tuning it, getting to the best customer outcomes.