Have you noticed lately that virtually every new technology innovation, system, or application is “powered” by artificial intelligence (AI)? Is it really? Hmmm. I guess it depends on how much you want to water down the definition of AI.
One of my favorite quotes about AI comes from British author and technology theorist Tom Chatfield who writes, “Forget artificial intelligence — in the brave new world of big data, it’s artificial idiocy we should be looking out for.” Amen to that.
I think maybe when we talk about AI what we should be talking about is “actual intelligence.” Especially when we’re using AI to talk about what technology systems can do. For example, let’s look at one of the features of Dell Boomi — Boomi Suggest.
When integration becomes a problem to solve, one of the tasks organizations want to streamline and speed up is data mapping. What data elements from applications A and B map to each other?
Because Dell Boomi offers a single-instance, multi-tenant iPaaS, it has many customers mapping the same applications over and over. Salesforce to NetSuite. Oracle HR to Salesforce. Oracle HR to Concur. Concur to Workday. You get the picture. And each mapping can be slightly different to meet a specific need.
The Boomi platform automatically logs these data mappings for all its customers. So, years ago a light bulb went off for Boomi. Why don’t we use these mappings — which number over 26 million now — to offer suggestions to our customers who are trying to figure out the best mapping scheme for their integration. And we can add a level of confidence to each suggestion to speed things up even more. Pretty intelligent. Actually.
Boomi Suggest is one of the reasons the Boomi platform is so easy to use. But the patented Boomi Suggest feature goes a step further in helping customers speed up integration.
Boomi Suggest is not just about data element to data element mapping. Quite often, especially with more complex data mappings where there could be 20, 30, or even hundreds of data elements that require mapping, there needs to be a data transformation before the data is mapped. We call the logic that controls this transformation a “mapping function.”
For example, if a company has offices in the United States and the United Kingdom, data transformation will be necessary when mapping certain data fields such as date format. In the U.S. we usually express dates in this form: month/day/year. The U.K. follows the European convention: day/month/year.
The straightforward way to do that is to write the logic as part of a mapping function. Before the value comes out of the source data mapping for date, the function executes a date conversion so the value corresponds to the European format when it is written to the target system.
Last time I checked, the Boomi Trust website, we had indexed nearly 115,000 mapping functions. And we use that information to make Boomi Suggest even smarter.
Boomi Suggest not only offers mapping suggestions with confidence rankings, it offers suggestions of mappings that include mapping functions. So, it’s very likely that when a customer needs a data mapping that includes certain transformation logic as well, we can provide it. It’s much easier to look at data mappings and mapping functions that have already been built — and add or subtract as necessary — than it is to start from scratch.
That’s a huge time saver. Actual intelligence.
Periodically, Boomi updates its AtomSphere platform. But, when we do, we want to make sure that our updates don’t “break” our customers’ integrations. So, we made it easy for them to submit their integration processes with sample data, which we can test against future releases of AtomSphere.
We call this optional feature Boomi Assure. When customers enable Assure, a Submit Test button appears when they’re in test mode. They put together a test package that includes integration process shapes, configurations, and sample data. After clicking the submit button, the package comes to us and we use it to test against all future releases.
Assure is a constant means of quality control for us. It’s also way for us to manage regression testing for a large customer population while giving them added confidence about their choice to go with Boomi. And, if we detect a problem — a process that fails — and cannot resolve within a release window, we will delay our release.
Just as Boomi Suggest benefits customers with the crowd-sourced knowledge inherent in millions of successful data mappings. Boomi Assure takes advantage of crowd sourcing as well.
Through Assure, Boomi tests thousands of integrations. The results of these tests benefit all our customers because it expands the number of data points we use to validate an upcoming release. Every time we run a test package, it subjects various combinations of shapes, data and integration logic to the discipline of regression testing.
Whenever a Boomi integration process runs, it produces log data. And whether the process executes successfully or fails, we store that data, including error messages.
For example, let’s say that someone has an integration process that runs every night at midnight. Then it fails. Why? Suppose the first thing the process does is connect to and get data from Salesforce, which it writes back to NetSuite.
Now when the integration process “wakes up” at midnight, it uses the Boomi connector for Salesforce with the access credentials provided by the customer. But, in this case, the password has expired so the integration process fails. That error message is captured by our connector and becomes part of the log for that customer.
The next morning when the customer logs into its process reporting, the status of this process shows red. It didn’t run. But they can click the Boomi Resolve button and Boomi will look at the error message and a pop up will say, “Your password expired.”
Of course, that’s a simple example. A process could fail from something much more complex. But, whatever the underlying issue, when you click Boomi Resolve, Boomi searches across the error messages from all our customers until it finds the best match. Then you’re connected to the documentation that will help you solve the problem in the fastest way, so you can get your business running again.
It’s the answer that best fits your situation, correlating error message data from thousands of customers. And it grows smarter the more customers we have, the more integration processes we execute, and the more correlations we make between errors and documentation.
Boomi Suggest, Assure, and Resolve. Actual intelligence that makes your business life faster, easier, and simpler.
To learn more about how Boomi can make your connected business run better, please contact our integration experts.
About the Author
Michael Morton is the chief technology officer for Dell Boomi.