No-code API tools for Excel wrangle data

Vault's journey to analyze at scale with no-code API

Vault, founded in 2017, is a relatively small company with 200 employees. Despite its modest size, Vault uses hundreds of spreadsheets to set insurance rates for its homeowner's, auto, collections and excess liability products. Initially offering its products in Florida, Vault's insurance offerings have expanded to most East Coast states, California and Texas with plans to expand nationwide. The company purchased a Coherent Spark license a year ago to enable a more rapid expansion of where its products could be offered.

While Spark's simplicity and portability factored into Vault's purchasing decision, it was the time factor that sold Vault on the software, said Robert Ross, CIO at Vault. Using Spark significantly reduced the time it took to manually convert actuarial spreadsheet data into an insurance product, he said.

"That process of going from a month and a half … down to a week was huge," Ross said.

Prior to purchasing a Spark license, Vault faced challenges with rapidly building out its infrastructure. Each of the four insurance products Vault offers -- homeowners, excess liability, auto and collections insurance for valuables -- has to be built for every state plus Washington, D.C., Ross said. The actuarial teams' process for determining rates and pricing resulted in 204 separate spreadsheets. Each spreadsheet could take more than a month to analyze.

Ross started looking for no-code API products that could convert Excel data on Vault's existing system into a service. Of prime importance was the flexibility for the tech to plug into Vault's existing system and continue to work as the company matured, he said.

Ross considered other options to wrangle its spreadsheets, including SpreadsheetWeb.

The full-stack, no-code product includes modules that customers can pick and choose from for different use cases, according to Ugur Kadakal, founder of SpreadsheetWeb and CEO of Pagos, which developed and distributes the tool. SpreadsheetWeb also includes a REST API interface where users can publish Excel files as generic APIs, Kadakal said.

Although the platform met Ross' needs, he didn't want a UI. "I wasn't looking for that much of the stack," he said.

In addition, Ross wanted the freedom to choose what app to connect to, when to use it and how to implement it, he said. Coherent Spark filled that need.

"I think [Coherent] did a really good job of focusing in on the problem and not trying to go too far to put lipstick on the whole thing," Ross said.

Spark reduced testing time

An additional bonus for Vault was that Spark created exact replicas of spreadsheets for Vault's system to use, Ross said.

Because the technology team didn't have to convert the spreadsheet into code on another platform, this capability removed the possibility of misinterpreting the spreadsheet's actuarial data, Ross said. This led to significantly less testing, he said.

"It was definitely nice to remove a lot of that [testing] out of our build timelines," he said.

Corrupted data is a common occurrence when converting spreadsheet data into app-usable data, said Holger Mueller, vice president and analyst at Constellation Research.

"The dirty secret of enterprise software is that users want to download data into spreadsheets, do their 'magic' and upload the results," which results in propagation of mistakes, Mueller said.

Coherent Spark provides the first step to automate the conversion of spreadsheet data into applications, Mueller said, which -- in addition to the scalability that Vault was looking for -- brings advantages of oversight, portability and security.