Using cognitive AI processes, CogniSure gives the insurance industry automated data structuring
“[The AI platform can convert] 100-page loss runs into something that an
underwriter or broker can make sense of within minutes without somebody manually keying.”
—Sai Raman
Founder and CEO
CogniSure
By Lori Widmer
It is no secret that the insurance industry is drowning in data. With over $859.9 billion in written premiums in the P-C market in 2023 (NAIC data), there’s a lot of time spent digging through paperwork. An Accenture study found that 40% of underwriters’ time is spent on non-core, administrative activities.
This kind of inefficiency is costing organizations plenty; Accenture figures place the loss of efficiency between $17 billion and $32 billion annually. Yet, organizations need to analyze loss runs, applications and other forms and documents in order to serve customers and compete in a tight market.
That’s where Sai Raman says his company can make a difference. Raman, founder and CEO of CogniSure, has employed cognitive AI—artificial intelligence that can emulate perception, learning, reasoning, and interacting. CogniSure’s AI platform can convert “100-page loss runs into something that an underwriter or broker can make sense of within minutes without somebody manually keying,” says Raman. “That’s what we do.”
He gives an example of a transportation account. There could be a CAB report, loss run, and driver schedule. Typically, he says, underwriters must look between schedules and loss runs to determine which drivers are having more claims.
There is a lot of time spent parsing data between various reports. What CogniSure does, Raman says, is “not just parsing out the data from a document, but stitching the data to make sense.”
Once the data is analyzed, it can be delivered to agency management systems, risk management systems, or any systems that carriers or brokers want, says Raman.
Applying tech solutions
That’s a big change, says Raman, who has spent much of his 27-year insurance industry career working on technology solutions to solve business problems. He started with a common issue: “I wanted to do a coverage gaps analysis in the policy” for the personal lines arena. That’s where he ran upon his first challenge. The only way to look at what coverage gaps existed, he says, was to read the policy documents.
With policy documents being in PDF format, Raman says he and his team figured out how to extract data from such documents. Once that milestone was achieved, Raman says he moved on to commercial lines. “Personal lines coverage gaps is a smaller problem compared to commercial lines with the loss runs and schedules. So, we pivoted toward commercial lines,” he says.
As part of the BrokerTech Ventures (BTV) Accelerator Program, Raman says he and his team got valuable insights and were able to hone their focus in the commercial lines market. “Talking to different brokers, we found out that more than the coverage gaps, these were bigger problems,” he recalls. “We were able to transition into solving the broker-centric problems with documents.”
CogniSure was formed in 2019. The platform officially launched in 2021. In the few years after launch and during the pandemic, Raman says he and his team worked within the BrokerTech Ventures Accelerator Program, “kind of refining our product market fit, working with a lot of BTV brokers during the 2020 COVID time.”
The time was well spent awaiting the reopening and resumption of businesses globally. “It was a great decision for us,” Raman says.
Now with over 15 customers and growing, Raman says the company is not resting. “We have expanded. In addition to P-C, we also have an employee benefits” solution to address the issues of self-funded accounts in the medical health benefits space.
“Employers are taking risks with self-funded plans,” he adds. “They need to know what’s happening in those plans.” Automating their documents works the same way that automating other insurance lines does—CogniSure can compile and analyze the data, reducing time, errors, and costs.
What users say
For Taruja Deshmukh, insurtech solutions manager for New Jersey-based broker Conner Strong & Buckelew, that kind of efficiency in her company’s employee benefits department was essential. “The data analytics team within our Employee Benefits group was spending a significant amount of time manually pulling unstructured loss data from various sources and preparing that data for further analysis in monthly reports provided to our clients with self-funded programs,” Deshmukh says.
“We needed a way to introduce automation to make this process more efficient, especially as the number of clients with self-funded programs was trending upwards,” she adds.
Tammy Brown, training and technology leader for Conner Strong, says, at first, the company tried working with an existing partner to solve the issue. “We attempted, unsuccessfully, to have an existing data warehouse partner build a functionality to create our monthly reports,” says Brown. “However, the process did not meet our expectations regarding scope and reporting flexibility.”
After considering other options, they settled on CogniSure. “Once we engaged CogniSure, it was evident their solution allowed for a greater repository of data allowing for real-time custom report generation,” Brown says.
It was an easy choice, says Desh-mukh. “Sai’s team had initially built a loss run insights platform for P-C, so we knew the underlying technology was already there.”
Deshmukh and her team worked with the CogniSure team to develop the employee benefits platform to handle the company’s loss data. “In doing so, we were able to develop a platform that cut down significantly on the amount of time the team spent on preparing client reports—from numerous hours to a matter of minutes to get the reports populated.”
The result has been exactly what Brown and Deshmukh had hoped for. Deshmukh says their data team is now able to spend more time on the higher-value functions, “assessing the data and delivering actionable insights to clients.
“The platform also provided a dynamic dashboard to better understand plan activity and cost-drivers and make data-informed decisions about employee benefit plan designs, including cost-saving opportunities.”
Deshmukh says that her team worked closely with CogniSure’s development team. “Over the course of several months we did a deep dive educating the CogniSure team on the current process, identifying data elements within employee benefit loss data materials and training the applicable carrier documents.”
It was a process made easier, says Brown, thanks to the willingness of the CogniSure team to work with them on the details of their processes. “Their team very quickly learned the nuances of our business both on the employee benefits and P-C side,” Brown says.
The end result is a process that Deshmukh says is a true collaboration. “Through the entire process, the CogniSure team was a great partner—their team was very flexible in adapting to our needs and committed to making a product that was beneficial to our team and the industry. And this applies to any work since then that we’ve done with their team.”
Conner Strong also gained in other ways. “Engaging CogniSure allowed our very talented data analytics team to focus on true analytic projects,” says Brown. “Prior to their engagement, much time was spent manually entering data into a spreadsheet with limited time to develop strategic solutions.
“We are now able to attract and retain talent that best utilizes their skills and abilities,” she adds. “Our clients have benefited from the high-end data support our teams can now provide.”
CogniSure has continued to lend support to the Conner Strong team. Deshmukh says she has used the CogniSure team’s expertise in other areas. “While we utilized a platform in our P-C claims group that supported automatic data feeds for a considerable amount of loss data, we still faced challenges with a manual process of converting loss runs that were not supported by automatic data feeds.”
CogniSure to the rescue. “By bringing in the CogniSure team on this problem, they were able to provide a product that helped fill a gap in our existing solution by creating an automated data conversion process for us to easily take loss runs in any format and convert it to a specific template to upload into our claims analytics platform,” Deshmukh says.
Even with customization, Raman says onboarding can happen within three to six months. Since the insurance-specific algorithms are already built in, he says customizing doesn’t create much of a delay in delivery of the product. “A carrier can go live within three to six months if they want specific requirements. Most of the cases, within three months [a customer] can go live.”
For Conner Strong’s tech team, their collaboration with CogniSure is something that has delivered quite a lot of value. “The CogniSure team is creative and forward thinking,” says Brown. “They understand the dynamic nature of our business. We have benefited tremendously from our partnership with them.”
The author
Lori Widmer is a Philadelphia-based writer and editor who specializes in insurance and risk management.
BrokerTech Spotlight offers a look at insurtech offerings from startups that have partnered with BrokerTech Ventures (BTV), a convening platform for brokers, carriers and wholesalers, and tech firms, and includes insight from other BTV member owners.
