AI’s key role in uncovering hidden value in document and data automation

This is the first Q&A in a periodic series highlighting women leaders in the fintech space. These interviews share valuable viewpoints on what it takes to bring a range of talent and skills to this fast-changing industry and what’s on the horizon for financial services and their technology partners.

A former attorney and now self-described serial entrepreneur, Shelby was named to Maclean’s 2024 list of the 100 Most Powerful Canadians, where she was one of 10 individuals in the artificial intelligence (AI) category. Other accolades include receiving Canada’s Most Powerful Women: Top 100 award, given by The Women’s Executive Network, and the Rising Stars award from Lexpert Magazine given to Canada’s leading lawyers under 40.

‎What was missing in financial services that your first, or second, or ensuing business ideas sought to solve?

My first foray into documentation was tilted toward legal technology. I was a partner at a law firm and watched thousands of documents being printed for trials or due diligence. I was inspired by the problem set and left the partnership to start my first business (applying data management techniques and new technology to large-scale documentation). This business grew to 400-500 people in about 4.5 years before being sold to Deloitte in 2014.

Arteria was first conceptualized while I was at Deloitte. We weren’t quite finished with the documentation space and found that the problem space in financial services was so rich. Documents in a financial institution are so critical – they are the workbench for a deal or transaction and underpin nearly every critical process. Moreover, it’s not just about getting the document done quickly (although that is critical). The data in documents is so critical, yet so underutilized, and there weren’t tools addressing the problem in that way.

To me, the solution is infrastructure that combines core documentation capabilities into one unified tech stack that is highly scalable across a large institution. This includes data intake, document generation, document digitization (i.e., structuring the data inside of documents), data outflow and intelligence. You must streamline the way data moves in between systems and documents and make sure those efficiencies are deployed at the heart of where the institution generates client value (i.e., trading, lending, onboarding, etc.).

Even today, do financial institutions generally understand how fragmented their own data is as they grow more interested in AI? What’s been the key to this conversation?

There is no question that strong data maturity is a catalyst for the effective and efficient deployment of AI. Since ChatGPT put foundation models into the hands of the world, organizations of all sizes have been under immense pressure to show value from AI. This has triggered significant investment into modernizing data management capabilities, which will elevate the business (in addition to streamlining AI deployments).

As with any transformation, business leaders should first look to identify areas where AI can drive outsized value. At the prioritization stage, velocity should be top of mind. That is, starting small, proving value and establishing momentum; these are the keys to success. In fact, starting with the hardest problems (i.e., areas where data is the most fragmented) can put the whole change program at risk.

The bottom line is every institution is working on better managing their data, but this is not necessarily a pre-requisite. Those who get to value the fastest will manage data work concurrently with core AI programs to prove value along the way.

Generative AI (gen AI) is fast expanding and requires some quick trust building given all the fascination, opportunity and trepidation. What is it about your own background, including as a recognized female disruptor and leader, that uniquely positions you to help make the case for this industry-changing technology?

Prior to Arteria, I ran AI (in addition to data and analytics) for Deloitte Canada. Significant advances in any technology, AI included, have triggered hype cycles where the market is fueled on potential. This can be highly lucrative to early players, but long-term value is contingent on tangibility – business leaders must be focused on specific areas where AI can drive real value.

Documentation is an area that is ripe for automation: approximately 90% of data in the enterprise is unstructured, and less than 1% of data is used in decision-making. AI has unlocked a whole wave of transformation in processes that are contingent on unstructured data. We believe we’re building the next big category of enterprise software and we’re very excited to lead our clients through the next wave of transformation.

Shelby Austin is CEO and co-founder of Arteria AI.