As seen on eSchool News

8 truths about K-12 IT systems management

Unique complexities can be distilled down to eight truths, and may explain why vendors never seem to meet expectations in K-12 IT.

By Gary Johnson
Director and VP, Utah Valley University and Kimono September 13th, 2017

Education administrators cope with so much data, it would be unfair to ask them to cope with even one byte more. Consider the information they handle every day. School districts in America today are complex, sophisticated businesses, not only managing multiple applications across multiple platforms, but also managing people and equipment in the real world, like bus fleets, library systems, and cafeterias.

Add it all up–including formative assessment systems, summative assessment systems, learning management systems, and special education systems–and you will find admins working with an average of 30 onsite and online platforms. That’s 30 systems to feed with data and update. The kicker is that those systems might not be on speaking terms with each other.

Interoperability is a multi-headed issue for any IT professional, but in the K-12 education world it is especially complex. These unique complexities can be distilled down to eight truths, and may explain why vendors who have been very successful in other IT verticals never seem to meet expectations in K-12.

Truth #1: The Solution Cannot Be Point-to-Point

With 30 applications to synchronize, choosing a point-to-point solution will land you in a hopeless tangle of data. Data from many active sources is profoundly difficult to keep current, especially when considering the different protocols used for each particular point-to-point integration. We’ve found that there needs to be a broker somewhere in the middle–one dashboard where you can see everything.

Truth #2: There Must Be Multiple Ways of Moving Data

Vendors are opening up their systems and allowing integration through RESTful APIs. Also consider the cooperation required to create the IMS Global Learning Consortium’s OneRoster specification–several vendors are part of that product steering committee. A successful broker/dashboard must be able to accommodate all of these integration methods. The broker must also be able to handle old standards, current standards, and emerging data standards. For example, OneRoster is looking like it will be the emerging roster standard. The broker needs to support it as well as the industry’s existing standards, such as SIF and CSV.

Truth #3: The System Must Merge Disparate Feeds

Data comes into educational systems from a variety of feeds, including CSVs and file sharing. Some of those legacy data inputs are just not going to go away. But that’s only part of the problem. The viable system broker must recognize that there no longer is a single, reliable data source. It must merge disparate data feeds, not just data coming from the SIS. Information might reside in an HR system, for example. Handling all these feeds develops a vital function, coveted by IT professionals and system admins everywhere: a comprehensive representation of the data truth of your district.

Truth #4: Your Data Solution Must Be Bidirectional

As classrooms go digital, rolling out more digital content and applications, the quantity of student performance data is increasing. Managing this data and securely sharing it back to a central system that can provide analytics capable of improving instructional approaches or aggregated results like report cards is one of the biggest headaches for educators today. Different systems don’t always talk to each other politely, and with some districts using as many as 30 applications, writing grades back to the SIS can get thorny.

Truth #5: We Need a Flexible Data Model

There are solutions now out there that are affordable, and some that are free. But the truth is that some of those free or low-cost integrations are profoundly rigid and can’t accommodate the data reality of school districts. For example, grades are not a typical data element that is supported by most roster data models. As a result, teachers spend time entering grade data, and that rigid low-cost or free integration creates a burden. A flexible data model is the answer.

Truth #6: We Must Deal with “Dumb” End Points

In the world of district data, we are moving toward REST APIs and other unintelligent end points. There is no inherent logic in an API that tells the system how to move data. And as mentioned earlier, many legacy systems still depend on CSV’s for data. The broker must be smart enough to gracefully handle not-so-smart end points. We need to put the brains–the application intelligence–into the broker or cloud platform.

Truth #7: Integration Belongs in the Cloud but Must Accommodate On-Premise Apps

That’s a rather long headline but it represents a complex problem. We’ve now experienced enough pain with alternative approaches to know the cloud actually is an ideal setting for interoperability, especially since so many of our applications are cloud-based. It gives you maximum visibility, maximum diagnostic capability and manageability. You can manage from anywhere, anytime.

But there’s a catch.

The truth of districts is that there are and there will continue to be many on premise applications, some of them legacy applications. These, too, must be integrated and managed into the system. The system must blend heaven and earth, bridging the cloud as well as on-premise.

Truth #8: Be Multi-Tenant with Supervisory Capability

For areas where intermediate units or a Board of Cooperative Educational Standards (BOCES) provide IT services to districts, the system admins need a big picture approach. The integration platform must allow the IU or BOCES to troubleshoot, diagnose, manage, and support multiple districts in one dashboard, but only show district personnel data belonging to their organization. State education agencies also have this need.

Summing It Up

The successful platform–one that is truly useful to district admins–is a broker solution that gets various applications talking, no matter where the data comes from. It’s a cloud-based solution because that is the most flexible, but it accommodates legacy systems that might be on-premise. It must be standards agnostic, able to use the preferred data standard and integration method for each connected application.

What we’re defining begins to look like an iPaaS (integration platform as a service).

There are several reputable companies that provide an iPaaS–in fact Gartner compared 20 of them in their 2017 Magic Quadrant for Enterprise Integration Platform as a Service. However, without a deep understanding of education data models, even these vendors may fall short, and may be expensive.

The team at Kimono has been working on the problem of K-12 interoperability for more than 15 years and have developed an iPaaS with an education-friendly price. Built specifically for schools and districts, they accounted for these truths and made sure it supports a growing number of industry-standard data models and protocols, such as OneRoster, LIS, LTI, ILP, SIF, APIs, and CSV.

About the Author:

Gary Johnson is the director of the Professional Sales Program at Utah Valley University, vice President at Kimono, and was formerly state account general manager at Pearson. He holds a Ph.D. from Brigham Young University in Instructional Psychology & Technology. His entire career has been involved with educational technology solutions.

eSchool Media uses cookies to improve your experience. Visit our Privacy Policy for more information.

Looking for some help?

Would you like to have an eSchool Media team member contact you?

Leave your details for more information