Daniel Dines

We gave up on some of our RPA stuff in order to come on the new technology, new frameworks, building from scratch, because we want to build an AI-first experience.

Look, we had a workflow engine, for instance, that was like Windows workflow engine. And we have perfected it over the years. But now we have switched to a more modern technology. that workflow engine, and we are building on the top of that. I resisted for so many years, and engineers in new IPath will realize how many discussions we had. Should we bring another workflow engine? Should we not?

Look, we had a workflow engine, for instance, that was like Windows workflow engine. And we have perfected it over the years. But now we have switched to a more modern technology. that workflow engine, and we are building on the top of that. I resisted for so many years, and engineers in new IPath will realize how many discussions we had. Should we bring another workflow engine? Should we not?

Look, we had a workflow engine, for instance, that was like Windows workflow engine. And we have perfected it over the years. But now we have switched to a more modern technology. that workflow engine, and we are building on the top of that. I resisted for so many years, and engineers in new IPath will realize how many discussions we had. Should we bring another workflow engine? Should we not?

AI finally convinced me that's the right moment where we should bring a new workflow engine that is specifically designed for what we call agentic orchestration. Specifically designed to facilitate very well the connection between agents and human users and other robots or other models, dedicated models and some other APIs, some other entry points.

AI finally convinced me that's the right moment where we should bring a new workflow engine that is specifically designed for what we call agentic orchestration. Specifically designed to facilitate very well the connection between agents and human users and other robots or other models, dedicated models and some other APIs, some other entry points.

AI finally convinced me that's the right moment where we should bring a new workflow engine that is specifically designed for what we call agentic orchestration. Specifically designed to facilitate very well the connection between agents and human users and other robots or other models, dedicated models and some other APIs, some other entry points.

Because I think very few people understand what are the use cases where RPA is really used and valuable, and why actually agentic doesn't work for those use cases. I can elaborate. Can we unpack them? So the sweet spot for RPA is to automate tasks that span multiple business systems and are of medium to high complexity. So they can span multiple steps. Usually it can be even 100, 200 of steps.

Because I think very few people understand what are the use cases where RPA is really used and valuable, and why actually agentic doesn't work for those use cases. I can elaborate. Can we unpack them? So the sweet spot for RPA is to automate tasks that span multiple business systems and are of medium to high complexity. So they can span multiple steps. Usually it can be even 100, 200 of steps.

Because I think very few people understand what are the use cases where RPA is really used and valuable, and why actually agentic doesn't work for those use cases. I can elaborate. Can we unpack them? So the sweet spot for RPA is to automate tasks that span multiple business systems and are of medium to high complexity. So they can span multiple steps. Usually it can be even 100, 200 of steps.

But the keys, they are rule-based. The input is structured and then the steps that you go are rule-based. But they, in a way, they capture the company knowledge. within the rules. Even simple things like if the VAT starts with this particular two numbers, then you have to take this particular flow. But you capture it in rules. This is very important. But these automations are very reliable.

But the keys, they are rule-based. The input is structured and then the steps that you go are rule-based. But they, in a way, they capture the company knowledge. within the rules. Even simple things like if the VAT starts with this particular two numbers, then you have to take this particular flow. But you capture it in rules. This is very important. But these automations are very reliable.

But the keys, they are rule-based. The input is structured and then the steps that you go are rule-based. But they, in a way, they capture the company knowledge. within the rules. Even simple things like if the VAT starts with this particular two numbers, then you have to take this particular flow. But you capture it in rules. This is very important. But these automations are very reliable.

They simply work until the underlying system changes. Now, when it comes to agentic, LLMs are actually not good at following repetitive steps. You are not going to have LLMs multiply to numbers. No, you are going to follow an algorithm and you are going to use creep language or you will program it, right? This is kind of the same with automations.

They simply work until the underlying system changes. Now, when it comes to agentic, LLMs are actually not good at following repetitive steps. You are not going to have LLMs multiply to numbers. No, you are going to follow an algorithm and you are going to use creep language or you will program it, right? This is kind of the same with automations.

They simply work until the underlying system changes. Now, when it comes to agentic, LLMs are actually not good at following repetitive steps. You are not going to have LLMs multiply to numbers. No, you are going to follow an algorithm and you are going to use creep language or you will program it, right? This is kind of the same with automations.

LLMs work relatively well when we are dealing with unstructured parts in a business process. It's sometimes the enterprise knowledge, it's difficult to express in rules. You can eventually. It's very difficult. There is a lot of tribal knowledge on the top of the public knowledge that human user is supposed to have.

LLMs work relatively well when we are dealing with unstructured parts in a business process. It's sometimes the enterprise knowledge, it's difficult to express in rules. You can eventually. It's very difficult. There is a lot of tribal knowledge on the top of the public knowledge that human user is supposed to have.

LLMs work relatively well when we are dealing with unstructured parts in a business process. It's sometimes the enterprise knowledge, it's difficult to express in rules. You can eventually. It's very difficult. There is a lot of tribal knowledge on the top of the public knowledge that human user is supposed to have.

So when you cannot express in rules, then you can build an agent that will mimic what the user will do, but with the intent that will reduce the human input on that part of the process. You cannot really eliminate a task using agentic. Because in a way, agentic AI, it's about delivering something autonomously.