Take APIs to the Next Level: Guidelines to Infuse AI into Your Business -- in Days!

Enterprises have been using APIs to connect cloud-based applications like Salesforce and NetSuite with legacy applications for a while now. Simon Peel, chief strategy officer at Jitterbit, shares insights on 2019’s trends to use APIs to connect to cloud-based AI-as-a-Service. 

Tags: AI, AI-as-a-Service, APIs, cloud, Jitterbit, iPaaS, machine learning, SaaS,

Simon Peel, Jitterbit
Simon Peel
Chief Strategy Officer

"By leveraging APIs, new AI-as-a-Service platforms offer the opportunity to apply pre-built AI products to existing business processes and create exciting new solutions."

Architecture Summit
Enterprise-Grade Integration Across Cloud and On-Premise
Online Conference

Enterprises have been using APIs to connect cloud-based applications like Salesforce and NetSuite with legacy applications for a while now. Today, there is growing interest in using those same techniques to bring artificial intelligence (AI) to the party.


New AI-as-a-Service platforms offer the opportunity to apply pre-built AI products to existing business processes and create exciting new solutions. These platforms offer public APIs that anyone can connect to. But what’s the most effective way to take these new APIs and apply them to your business where they make the most sense?


This article explains how to leverage API Integration Platforms to quickly reap the benefits of AI-as-a-Service in any business.


Artificial Intelligence: It’s Rapidly Becoming a “Buy Versus Build” World

There are over 4,000 AI providers, and that number is constantly growing. Many of these services are accessible via publicly exposed APIs. That means you can start using AI with very little manpower, and at very little cost.


When AI was still an emerging technology, enterprises had to build their own AI-powered applications in-house. Teams of developers and data scientists would use machine-learning (ML) tools to build and train AI platforms to look for business insights or automatically perform business tasks that were previously handled manually.


But in today’s technology market, there’s a plethora of AI-as-a-Service providers with offerings for every industry, from financial services to healthcare to HR. With thousands of AI services available off-the-shelf that provide proven business value, it makes sense to find a third-party AI service and infuse it into your existing business processes.


These AI-as-a-Service offerings are similar to the cloud-based apps that enterprises increasingly rely on in today’s “best-of-breed” technology market. In the same way that the APIs for those popular SaaS applications, like Salesforce, empowers companies to connect on-premises systems and data with cloud technologies, AI-as-a-Service applications also come with publicly exposed APIs that can be connected to a broad range of business processes. That means that companies can essentially follow the same path they have used to leverage best-of-breed cloud apps and immediately leverage the power of AI.


4 Guidelines for Using APIs to Infuse AI into Business Processes

The proliferation of third-party AI services with publicly exposed APIs presents a major new opportunity for businesses to take advantage of cutting-edge smart technologies such as sentiment analysis or image recognition through AI platforms like Google AI or IBM Watson.


But leveraging this technology to produce real value still requires the right strategies and tools. Just like with the popular cloud-based apps, connecting to APIs for AI services isn’t automatic.


Here are several guidelines that will empower your company to quickly infuse AI into your business processes:


Guideline #1 -  Figure out early on where AI really fits. One of the crucial steps happens very early ideally in the planning stages when you are collecting requirements and assessing the needs of various stakeholders – identify where AI will help you the most. 


Try to really understand what problems AI can solve for you and the business value those solutions will bring. Where does AI naturally fit in?  Who needs to use the AI tool directly, versus benefit from its “ah-ha” moments inside their existing tools?  It will help if you don’t think of AI as only a standalone tool, but rather as something that will also augment your existing applications and processes with knowledge. Where will that one AI-driven insight be best placed in your existing apps so that users get the biggest bang for their buck at the point of action?


Guideline #2  - Accelerate AI projects using an API Integration Platform. All AI solutions come with pre-built APIs that can be used to inject data and gain access to insights from the AI platform. But how will your existing data and applications be connected to those APIs? Or put another way, how will the AI system tell people it has discovered that “ah-ha” moment?

Who will deal with the significant differences in accessing APIs and data formats in the various systems? What about the logic that needs to be applied between systems to decide where these learnings should be surfaced, in which applications, with what level of security applied?


The answer:  Businesses can use an API Integration Platform to connect AI APIs with their existing business apps, transform the various data types automatically so they match exactly what the source/target systems are expecting, and apply simple or complex business rules using an easy-to-use graphical interface.  Most importantly, all of this can be done in hours/days versus weeks/months. 


Guideline #3 - Auto-create APIs. What about existing applications and data that don’t already have APIs?  Use an API Integration Platform to create new APIs automatically, account for security concerns, and auto-generate the relevant documentation for internal or external stakeholders in a developer portal, so it’s clear how to use the new APIs.  For example, with a few clicks, a business can auto-create an API to front-end their proprietary bug-tracking system so any customer can ping the API and get the latest status of their defect resolution, 24 hours a day, without placing a single call to the helpdesk.


Guideline #4 - Encourage experimentation. Many AI projects won’t show business value right out of the gate. Enterprises need to provide room for experimentation and enough runway to allow projects to get off the ground. One way to help ensure AI projects succeed is to gather the right tools and support ahead of time. With the right tools, companies can accelerate the delivery of AI projects and give users the room to try different things without expending much manpower or budget.


Bake privacy and security into your products. Privacy and security are just as important as ROI metrics when infusing AI into business processes that handle enterprise and customer data. Establish rules and processes around both from the very beginning of your development cycle.


The great news is that today’s API Integration Platforms can make it easier to maintain uniform policies around data exposure. But check to be sure that these platforms provide a holistic approach to API and integration. To experiment safely you should avoid solutions that patch together point products for different tasks and don’t provide uniform data governance.


Every AI Journey Starts with a Single ‘Smart’ Step

AI is already being used in some exciting ways, but adoption will accelerate dramatically now that we’ve entered the era of AI-as-a-Service, especially if you keep in mind that creating new business processes infused with AI does not require you to start from scratch.


By leveraging API Integration Platforms and AI-as-a-Service APIs, you can compose new solutions by augmenting existing business processes in days versus months or years. There are countless opportunities to leverage AI infusion and the technology to do so is well within reach - you simply have to take the first step.


As Chief Strategy Officer and CMO, Simon Peel leads global strategy and marketing for Jitterbit, creator of the Jitterbit API integration platform for quickly connecting apps (SaaS, on-premise, and cloud) with intelligence into any business process.  Simon believes in finding growth strategies where no one else is looking, which accounts in part for his interest in exploring the marriage of APIs and AI-as-a-Service.