Erik Reeves, chief technology officer at Boston-based Anaqua, has seen a great deal of change in the Software-as-a-Service space over the last several years. He spoke with StrategicCIO360 about what’s shifting, particularly for Anaqua’s IP clients, and how to make AI work for your organization.
What trends have you witnessed among SaaS providers in the last three to five years?
We have seen many interesting global and economic forces in the last few years impacting those in the SaaS space. The pandemic—and all the natural consequences—challenged every organization across industries. I have observed some trends specific to the intellectual property field, which is where many of Anaqua’s customers concentrate. A few of those trends include a rise in the connected IP ecosystem and the global, remote workforce.
Like many organizational systems, IP is very cross-functional and cross-disciplinary. It has also traditionally operated without a unified leadership structure until the last 20 years, so its maturity as a “corporate function” is a bit less mature than some others. The “connected IP ecosystem” is the combination of software, services and necessary integrations that make up all aspects of the entire IP lifecycle, including the use of cloud technologies that facilitate the overall end-to-end experience.
Looking at the continuum from early innovation capture and scouting through formalizing IP to filing patents or trademarks, to managing the prosecution, and maintaining the IP assets through to final disposition is conceptually uncomplicated.
From an IT perspective, there are many connected systems that interact with this core process, including patent and trademark offices, document management systems, financial systems, outside counsel (or in the case of law firms, their clients), foreign filing services, reporting systems, etc. Companies and firms that can integrate and establish connections between systems are going to see efficiency and effectiveness.
There is a need for an “open” ecosystem from an interoperability standpoint. Law firms, corporations and service providers cannot ignore the colossal opportunities that have emerged because of the cloud revolution. The services, capabilities and the pace of new productization must be part of the mix as companies look at their evolving digitization strategy.
Artificial intelligence has been a trending topic for a while. What are your views on automation and on its workflow applications?
The promise of AI has been widely discussed since the 1950s, but in the last 10 years, those discussions have evolved to talking about AI’s abilities to solve business and social challenges in more concrete ways.
AI has practical applications in environments with consistent and homogenous data sets, like image recognition, however, where data is messy, incomplete and complex, it is harder to see a clear application. That being said, AI is often glorified as a futuristic solution that can do it all—everyone is using or developing AI solutions these days. But I think that’s the wrong approach.
Organizations sometimes fail to think about how to invest intelligently and apply the right tools to the right problems in AI and ML. The most successful applications of AI are when teams step back and identify a problem that is hard to solve simply and lends itself well to today’s AI applications.
I’ve also seen that some of the best and most impactful use cases of AI, particularly related to workflow applications, are those that automate repetitive, mundane or even annoying tasks.
Further, good AI relies on good candidate data. Just like valid statistical analysis requires certain approaches—or correcting mechanisms—to data and population sampling, useful application of AI requires data that meets certain reasonable criteria for quality, size and distribution. Without pointing fingers, it is not hard to find poor, or just wasteful, uses of AI.
AI projects can be expensive, so ROI has to be top of mind. Organizations should focus on the problems, assess which tools can help them, and be mindful of the powerful capabilities of AI in this context, but not “start with the hammer and search for nails.”
How can AI create demonstrable differentiation?
Plenty of “successful” projects ultimately fail because there simply wasn’t sufficient differentiation from a vastly simpler and cheaper approach. The results may have been better, but no one cared enough about a tiny improvement with so much added cost and complexity—you need to find leverage.
Conducting large-scale document data processing, and image processing are two valuable applications of AI that I have seen. AI tools can help enhance search capabilities or techniques to make suggestions for improving queries and search results. Even aiding in the generation of legal arguments or responses to things like patent office actions are other areas worthy of consideration. Using advanced tools on document processing, optical character recognition and intelligent data extraction can improve data quality, decision-making and ultimately ROI.
Starting from scratch rarely makes sense with an AI/ML investment. The leading cloud providers have already invested countless millions into R&D in this area to deliver new products and services every year, and enterprises should take advantage of these. Before looking at building a new capability, service or infrastructure, one should absolutely evaluate if there are cloud services that can provide it or be adapted to meet the demand before investing and taking on added risk and cost.