Versori is transforming into an Agentic iPaaS by integrating AI agents into its platform to simplify and accelerate the process of building, testing, and modifying integrations, with Jay, their central AI solution architect, driving the vision for scalable, efficient, and cost-effective integration solutions.
With the rapid growth of AI and LLM technology, we knew here at Versori that becoming an Agentic iPaaS would be the single biggest shift in our trajectory. Therefore in early 2024, we set about building AI into our architecture. For our Agentic Platform we have embedded AI agents into the foundation of each function and tool of the Platform’s integration capabilities. Each agent functions independently of each other and together as one with Jay, our central AI solution architect agent employing each micro agent as an SME to plan, build, and validate any Integration between any number of systems.
Development of Jay began in early 2024 as a lightweight prototype. We wanted to test our basic premise and embed AI into our existing technical stack as quickly as we could in order to validate our assumptions. This meant leveraging the Go programming language for some of the implementation.
Shifting to present architecture, the full AI stack is written in Python with multiple agents across multiple functions. By moving the dependency on Cloud Eventing to the Frontend we were able to remove any reliance on Go in our AI micro service and put ourselves into a great position to flexibly scale our AI using Python.
As touched on previously, for performance, and flexibility we use multiple agents, each dedicated to a particular task. By designating specific agents for both micro and macro functions, which are orchestrated by Jay, we have significantly minimised the common pitfalls often encountered with AI agents. From this change we have seen increased accuracy and incredible growth and capability.
Building integrations can often feel like solving a puzzle. You start with a clear idea of what needs to be done, but translating that high-level requirement into a functional integration is where the real challenge lies. That’s where Jay steps in, simplifying the process from start to finish.
Defining the Integration Requirements
Jay removes the complexity of turning high-level requirements into actionable steps. Users provide a simple prompt, for example; – ‘When a reservation is made in Booking.com, create a new reservation in Hostaway and if the guest stays no longer than 3 nights create a customer record in Netsuite’ – and Jay interprets it, eliminating the need for manual planning or setup.
.png)
Building the Integration Flow
Jay constructs the integration directly, mapping out workflows and configurations on the board, reducing the dependency on a user having detailed knowledge of how to build the Integration.
Seamless Testing and Validation
Testing is built into the process. Jay identifies what needs validation, runs tests seamlessly, and highlights areas requiring attention. If adjustments are necessary, Jay guides users through the exact changes needed to ensure the integration works as expected.
Real-Time Modifications
When requirements change, modifying an integration is simple. Users can update their needs, and Jay adapts the integration live, making real-time adjustments without disruption.
From a product perspective, it was important for Jay to feel integral to using the Platform and building Integrations. Rather than AI and Jay feeling secondary to a user’s experience, our goal was to make Jay central to the integration-build experience. This vision is reflected in our latest UI update, which offers a much more immersive experience, with all planning, building, testing and deploying done centrally using Jay.
Reflecting on the past year while looking ahead to the future, it’s clear that the meteoric rise of AI has transformed the iPaaS landscape and redefined the technological solutions businesses must provide. At Versori, we aim to lead this pivotal shift by becoming the premier Agentic iPaaS platform. Our mission is to empower customers with AI to build Integrations with ease, speed, and cost-effectiveness, accelerating their time to value to extraordinary levels.
