Recently I had the chance to meet with co-founder and CEO Pat Utz of Abstract, an AI startup focused on uncovering the impact of government and regulatory shifts across a multitude of stakeholders including government entities (city, state and federal) enterprises both small and large, as well as to the public. Abstract’s overriding goal is to streamline the workflow of public policy analysts, government/regulatory affairs professionals, lobbyists, and law firms , by delivering AI-driven insights that are created by its ability to ingest, connect and predict policy impact
Since its founding, Abstract has worked with over 60 organizations ranging from cities, non-profits, colleges to Fortune 500 companies. The startup has also attracted a number of early investors including Vitalize VC, Amplify.LA, Motivate Ventures, and Stage Venture Partners. Advisors include Facebook’s Director of Public Policy Sarah Wynn Williams, Former Data Scientist for Lyft Rex McArthur, Global Public Policy Lead for Block Bijan Mehryar, and Former EVP, Chief Legal and Risk Officer for HITRUST Kimberly Gold.
Here is my conversation with Pat.
M.R. Rangaswami: How did you become interested in improving government transparency and building a better way to track and analyze information about national, state, and local policy?
Pat Utz: All three of us founders grew up in families connected to politics and policy. My father ran for office in Argentina and later came to the US to learn English so that he could have a better shot at working in international policy in Argentina. Mohammed (Mo) Hayat’s family has a long lineage of government ministers in Kuwait. And Matthew Chang’s family hails from Taiwan, specifically the government’s Ministry of Economic Affairs.
The three of us grew up with discussions on geopolitics at the dinner table. For example, I heard the wrenching stories of how families in Argentina were affected by failed government policies that led to hyperinflation and corruption across government and business. In college, when discussing pending legislation with friends and fellow students I was dumbfounded that no one had actually read the proposed legislation – which of course led to misinformation and confusion.
So, when Mo and I were deciding on our senior Electrical-Computer Engineering thesis at Loyola Marymount University, we came up with the idea to use natural language processing algorithms to make “abstracts” of legislative bills. We not only had our thesis topic, but also the name of our soon-to-be startup. The mission to increase government transparency for the public good was born.
M.R.: Can you give us a sense of how technology can be used to help manage the increasing complexity in the legislative space?
Pat: Absolutely. In 1970, there were approximately 400,000 restrictions in the Code of Federal Regulations and by 2019 that number had swelled to one million according to QuantGovt. The Office of the Federal Register estimates that there are 3,000 to 4,500 final rules published each year. Keep in mind, there are over 145,000 governing entities now in the US – cities, counties, states, and federal entities. In California alone, over two million changes a year happen at the state level across all the regulatory agencies. And as far as the bottom line, today the average US firm spends “between 1.3 and 3.3 percent of its total wage bill on regulatory compliance,” according to Francesco Trebbi and Miao Ben Zhang in “The Cost of Regulatory Compliance in the United States.”
The last two decades have seen significant shifts in managing and analyzing all this data.The first shift occurred when the data went from paper to digital, and everything became available on the web. This spawned the first generation of government legal tech software solutions, and with the advent of social media, policy teams added media monitoring tools like Meltwater and MuckRack to their arsenal.
The second shift happening now involves using AI not just to track everything, but to actually make sense of it. This includes evaluating the actual language used, then analyzing news on proposed legislation from every channel, coupled with official and “citizen rumble” commentary to understand things like which politicians and groups favor which agendas, and who is influencing them. All of this needs to be abstracted to create a cohesive “what’s next” plan: that is where Abstract fits in. With AI, we can interconnect all the noise to actually infer what’s relevant to a stakeholder, be it a company, city, state or any government agency.
M.R.: Everyone today is talking about AI, not just for writing better emails with Generative AI, but considering how purpose-built AI models can to solve specific challenges. What do you see as the role of AI in the policy world?
Pat: We believe that AI can level the playing field for policy making since it can distill the noise around government regulations, and then enable stakeholders to quickly spot where the risk and opportunity lies. It used to be that only large teams with deep pockets could do this, but with AI, organizations of all sizes from small non-profit organizations, to large cities, to multinational corporations can use tools like Abstract to dramatically reduce the costs of their legal and government affairs efforts. On the flip side, this gives policymakers better insight to make more equitable and more informed decisions.
Until now, the actual analysis of all of this data and shaping of policy has largely been left to analysts generally fresh out of college and to lobbyists who are typically working with special interest groups to influence legislation. The proposed legislation is then voted up or down by elected officials who often have little understanding of the issues the proposed policy contains or how it will affect constituents. For many issues, including the future of AI itself, this can have profound consequences for millions.
One of our chief goals with Abstract, is to become the end to end solution for understanding how any change or existing government policy impacts your business, whether it’s a risk in fines and costs or an opportunity to create more revenue. I envision a future wherein presidents across the world ask Abstract for macro-policy advice.
M.R. Rangaswami is the Co-Founder of Sandhill.com