iStock.com / phonlamaiphoto
Pamela Negosanti, global vice president of insurance at Expert System, explains how the company assists clients in a world of increasingly high volumes of information. She will be speaking at the Intelligent Automation & AI in Insurance Europe event in London on May 21.
“Underwriting is a sweet spot for natural language processing and natural language understanding.”
Tell us about Expert System.
We help enterprises to become more competitive in a world that requires ever-faster processing of increasingly diverse, high volume information.
We do this through artificial intelligence (AI). Expert System provides an accurate, automatic and immediate understanding of text—in other words, our technology and applications mimic human comprehension, but in a fraction of time.
What kind of companies do you work with, and how do you help them?
Our solutions have been deployed across many industries including banking & insurance, government, media & publishing, life sciences & pharma and oil & gas.
In particular, we’ve been working with some of the world’s largest insurance and financial institutions to:
- Automate customer service: chatbots, natural language Q&A, automatic email classification and intelligent search applications;
- Enhance AI robotic process automation for: loan approvals, claims management, policy underwriting, audit support and more;
- Mitigate operational risk: third party risk management, anti-money laundering and legal compliance procedures, cybersecurity intelligence;
- Enrich customer analytics: analyse and predict consumer behaviour and trends in real time to support audience targeting; and
Augment knowledge management: support corporate intelligence, investment strategies and competitive activities.
How is AI transforming insurance?
AI is transforming the entire insurance industry—in less than 10 years the industry will be completely different. AI is impacting the entire insurance value chain from product design to distribution, underwriting and claims.
The real transformation is already feasible from a technical perspective; the hurdles are related to adoption and organisations’ ability to embrace change and innovate.
Why is underwriting the perfect match for natural language processing and understanding?
The capability to understand deeply is a key factor for several reasons. Underwriting is strongly based on documents, which typically are:
- Diverse in formats and content; and
- Rich in intra- and cross-references.
As a consequence, unless you truly understand the documents (policies, reports, external sources, etc) you cannot come to a conclusion, take a decision or an action. This is the reason why underwriting is a sweet spot for natural language processing (NLP) and natural language understanding (NLU).
What underwriting activities can these technologies be applied to?
They can be applied to a variety of activities and processes, such as:
- Submission (speeding up the submission process by automating the necessary data to make an offer);
- Contract comparison (comparing different policies such as master vs local, or slips, renewal);
- Risk assessment (analysis of third party risk reports to detect risk factors and grade them);
- Clauses similarity (checking the risk exposure when it comes to how clauses are phrased);
- From unstructured policies to structured policies (detecting the general information, the coverages and exclusion of manuscripted policies); and
- Cognitive underwriting advisor (providing support to underwriters by collecting information from different sources, both internal and external).
How can this technology improve risk exposure reduction, and augmented sales capacity and timing?
Risk exposure reduction: today, risk assessment relies primarily on humans. This limits the amount of data and documents that can be processed, and exposes results to a certain amount of subjectivity.
Technologies such as Cogito, Expert System’s AI platform, can process the data and documents as accurately as a risk engineer, with objectivity, in a fraction of time.
Augment sales capacity and timing: the productivity gains provided by AI will empower the entire workforce. Think of the submissions process and requests for quotation. With this technology, the number of offers an underwriting team can make will be multiplied, all the while dramatically reducing processing time. We all know that presenting an offer first is definitely a competitive advantage.
How can this interact with and leverage big data?
There are perhaps only 10 companies in the world which have big data. Even if they produce a lot of data and documents, insurers and reinsurers aren’t really in this category. There is no one-size-fits-all technology. A lot of companies have experimented with pure machine learning techniques, which if applied to complex cases may reveal that they are insufficient to become productive, with approximately 60 percent accuracy.
Depending on the use case and scenario, a combination of approaches or a hybrid approach could be a viable solution.
How important will data be to the insurance industry?
More than ever, data gives insurers a competitive advantage by enabling them to grow their business, by (i) increasing sales capacity and reach; (ii) improving the customer experience, hence the retention; and (iii) gaining efficiency.
Insurers who are late in adopting AI will ultimately struggle to compete and lose market share.
Pamela will be joined on stage by 40+ insurance trailblazers ready to deliver detailed case studies on how to leverage AI for commercial lines.
To secure one of the last few remaining tickets for Intelligent Automation & AI In Insurance Europe (May 21, London), register here today:www.intelligentautomationeurope.com
Expert System, AI, Intelligent Automation & AI in Insurance Europe event, Banking, Insurance, government, media & publishing, life sciences, pharma , oil