The new era of integration: Celigo’s AI platform features

Imagine IT and business teams working in lock-step to build integrations, developers focusing on creativity not technical complexities, and errors fixing themselves before you even know an issue occurred. That reality is closer than you think.

The recent strides in AI (artificial intelligence) have fueled our mission of turning every iPaaS (Integration Platform as a Service) user into an integration expert.

With AI as the driving force behind it, our iPaaS enhances building and error management, bridges knowledge gaps between IT and business users, and provides insights learned from thousands of customers. And we’re not stopping there. We are continually improving our platform capabilities and working towards that vision of perfectly in sync teams empowered by simplified integration.

Let’s explore the AI capabilities already built into the Celigo platform and where we’re headed next.

Enhancing efficiency

When approaching how to enable our platform with AI, we started with areas where our customers were facing the greatest inefficiencies: error management and building integrations–specifically around writing code.

We built an error classifier using AI and natural language processing to better categorize errors. From there, we implemented a mechanism to automatically resolve integration run-time errors from external APIs–resulting in 95% of errors being automatically resolved.

When it comes to writing code, we began with JavaScript and handlebars. The goal of this feature is to allow users to describe the problem they’re trying to solve and have AI produce the code for them.

 

Watch this demo of how Celigo’s AI produces JavaScript.

These capabilities make using our platform more efficient for customers and help accelerate the integration process.

Bridging knowledge gaps

Democratizing the integration process ensures team alignment, prevents bottlenecks, and drives innovation amongst other benefits. One of the biggest challenges to achieving this is handoffs between technical and non-technical users.

AI allows non-technical users to better understand the integration process. Integration isn’t always transparent, so our AI acts as a translator. It helps users understand what their automations and configurations mean. Our chat-enabled editors that support natural language and our auto-documentation of existing flows contribute to streamlining this process.

 

Explore Celigo’s new capability to provide a description of each flow.

Creating connected experiences

Our AI continuously learns and adapts, leveraging past user successes to enhance integration and ensure a better experience for future users.

For example, standard integrations can be replicated across thousands of customers. This allows us to empower all our users to learn from others. To help with this, we created a knowledge navigator that leverages internal and public data sources and embeddings to power the Celigo AI Knowledge Bot with detailed knowledge of our product.

 

See how Celigo’s knowledge navigator helps explain how to use the platform.

What’s next

With the in-house experience we have gained since we began implementing AI in 2019, we’re continuing to evaluate new ways to use it to improve the user experience.

We leverage AI internally to read API guides, assist in comprehending the docs, and extract the metadata. This has allowed us to manage ongoing maintenance of API connections at scale.

We are also doing our due diligence to ensure the functionality we rollout is trustworthy. By verifying safety, privacy, and accuracy, we can ensure our tools are reliable.

Next, we are looking to develop a recommendation engine with a custom deep learning model. This will help guide users through their entire integration journey.

Dive deeper into our new AI capabilities in our latest on-demand webinar, “Transforming iPaaS with AI.”