AntWorks uses artificial intelligence to process documents for a range of industries, including insurance. InsTech’s Ali Smedley spoke with the global head of AntWorks’ insurance business, Neil White, to discuss the differences between OCR and AntWorks’ CMR+ offering, how insurers can use it and what the future of data transformation will look like.
Ali: What does AntWorks offer to the insurance industry?
Neil: AntWorks is an intelligent document processing platform. We have clients across a range of industries including insurance. Insurers and brokers can use AntWorks to process unstructured data from complex documents such as policies, slips and quotes to increase efficiency, unlock data and generate market intelligence. For example, extracting data from policies can be a big challenge for brokers. This is because various insurers will use different formats, structures and terminology. This is where AntWorks comes in.
Ali: Neil, what is your role at AntWorks?
Neil: I’ve been in the technology sector for 25 years and now work as Head of Insurance Practice at AntWorks. The company wasn’t originally looking to focus on the insurance industry, but it started to attract insurance customers because of the data extraction challenges that the industry faces. As a result of that interest AntWorks built up its insurance-specific expertise, including hiring people who previously worked at major insurance companies such as Allianz and Zurich. This was to ensure that the solution was developed and tailored to meet the needs of the insurance industry.
Ali: AntWorks offers Cognitive Machine Reading (CMR). How does this differ from OCR?
Neil: OCR (optical character recognition) can work well for extracting data from simple structured documents, but when the documents are more complex, there are variations in the formats or where contextual understanding is required, it becomes much more challenging. AntWorks offers CMR as an alternative to OCR. CMR learns from the data and can recognise patterns and sentiment. For example, the UK and the US may use different words to mean the same thing, such as professional indemnity insurance in the UK and errors and omissions insurance in the US. CMR has learned that these are the same, but OCR does not have this capability.
AntWorks was profiled in the InsTech report Data Extraction and Ingestion: The 40 companies to watch in 2021. The report examines the challenges of working with unstructured data - it is available to download here.
Ali: What data can AntWorks’ technology ingest?
Neil: Antworks is able to ingest typed writing, signatures, stamps, barcodes and handwriting. For example, this includes cursive handwriting that could be on an automobile claim form. We do not provide audio, video or image recognition, but Antworks can extract information from anything else that may be on an insurance-related form or document.
Ali: Can you give some examples of where AntWorks fits into the insurance workflow?
Neil: Within the industry, commercial insurance is where our focus is. AntWorks can assist with the placement stage by extracting or comparing data from slips, policies and quotes. For the claims process, it can extract information from forms, invoices or loss reports and understand contextual information about the claim. Claims and policy related documents are highly unstructured and difficult to ingest. AntWorks can extract data from these documents which can help insurers and brokers utilise the data effectively for analytics, pricing or predictive purposes.
Ali: What do you think the future of data transformation will look like?
Neil: The automation of data extraction isn’t about replacing people: it’s about letting people use their skill sets better. With automation it becomes cost-effective to capture data that would be prohibitively expensive to gather manually. At the same time it allows experienced underwriters and brokers to spend more time doing what they do best: serving customers and adding value. In the future, more processes are going to become digitised and automated, but there are challenges with dealing with legacy systems and regional challenges that need to be overcome. The data extracted by solutions such as AntWorks will be increasingly used for data analytics and decision making.
Ali: Why has AntWorks joined InsTech as a corporate member?
Neil: Being part of the InsTech community helps us better understand how to make sure our solution really is just that; a solution to the very real challenges insurers face. At the same time InsTech also helps us better communicate what we have to offer. We’re already seeing Antworks technology change the way insurance companies handle documents and the data they contain. InsTech really helps us keep our finger on the industry pulse.
Ali: What companies is AntWorks looking to connect with?
Neil: AntWorks is ready to help any company within the insurance market. Any enterprise with hundreds of thousands or even tens of millions of documents to process could feel the benefits. Our current clients include banking, finance and investment institutions, professional services companies, technology providers and business process outsourcers.
For more information on AntWorks, go to www.ant.work.com