AI Meets Law: Streamlining Patent Drafting with Smart Automation (Jack Xu of PatentPal)
Jack discusses pursuing entrepreneurship with a legal background, being a solo founder and transforming patent law with technology.
About
Welcome to the 24th installment of Foothill Ventures’ Lessons from Founders series. Every week, we publish an in-depth founder interview, ranging from early-stage entrepreneurs to successful businesses. Our conversations cover their personal journeys, the lessons that shaped them, their visions for the future, and their failures. We also learn more about their companies and about the challenges they try to solve. These insights and lessons are applicable to any entrepreneur — current or future.
Read past interviews here.
PatentPal is a team of attorney-engineers seeking to improve the experience of drafting patent applications. Specifically, our software automates much of the mechanical writing and repetition across claims, diagrams, and description sections of the patent. You can try our product at www.patentpal.com or reach us directly at info@patentpal.com.
Jack Xu, the founder of PatentPal, combines his expertise in patent law and machine learning to build language generation software for automating patent drafting. Prior to founding PatentPal, Jack worked as a patent attorney at globally-recognized law firms, specializing in artificial intelligence and software patents. Jack holds a JD from Harvard Law School and a BS in computer science from the University of British Columbia.
Why we invested in PatentPal: We have looked at legal tech for several years, searching for a deal that met our bar. We believe that legal services will be transformed by NLP/ AI, and that patent drafting is an obvious area — the drafting is highly structured, and a knowledge graph approach can create huge value. This is primarily a bet on a founder: Jack Xu is a talented computer scientist who is also a talented attorney, and that he has the scrappiness to make initial headway with customers and product development. Beyond the bet on the founder, this is a bet on data: the software will learn from multiple drafts of patents, and will develop a patent knowledge graph that should improve with time. The value of patents continues to grow, and law firms increasingly must turn to automated tools to efficiently draft and file: we want to be part of a learning system that gets off to an early start.
Jack introduces himself
I’m building a company called PatentPal. We are using natural language processing technology to build products that assist with drafting patent applications. Much of our work is at the intersection of machine learning, computational linguistics, and patent law. Our goal is to automate workflows in creating new intellectual property assets.
His early background: early exposure to business and tech
Two aspects of my early experiences stand out. First is coming from a business family where we talked about work all the time — at the dinner table, etc. I didn’t realize that this was not normal until much later. Second is being exposed to software at an early age. Our family owned a computer as early as my first year of primary school. I wrote my first computer program at age 11: it was a video game that taught kids new English vocabulary. I had a lot of fun making it — it won awards and prize money at our provincial science fair in Canada. Looking back, these were happy accidents that exposed me to the world of business and computers early on.
On pursuing law school and patent law
I wanted to do something wildly different. We didn’t have any lawyers in our family and most of my training had been in computer science. I was interested in learning to see the world through a legal lens. And more generally, if something scared me, I had to do it.
At the same time, starting a company was something I knew I would want to do. So, I asked myself what skills I would want to bring into my role as a founder. While I think soft skills are more important for working in a team or organization, I preferred getting degrees in hard skills. I felt that a law degree would give me the credentials to practice a profession while also teaching me about business.
If something scared me, I had to do it.
Patent law was something I discovered during law school. It aligns with my technical background — you cannot practice patent prosecution without required coursework in science and engineering. It dealt with fascinating synergies between law and innovation, like applying property rights to ideas.
On practicing patent law & drafting a patent for a cardboard box
Patent law is not something that most people regularly come across. But in Silicon Valley, patents play a major role: the two most dominant practices at major law firms here are corporate and patent law. Working as a patent attorney was fascinating because you work with different technology every week, and always the most cutting-edge inventions. I had a particular focus on artificial intelligence, but I also drafted patents directed to computer hardware, robotics, telecommunications, and even a new cardboard box design once. My work gave me a high-level view on the innovation that’s happening in Silicon Valley and the tech industry in general. It also inspired me to think about how I could apply the latest developments in machine learning and natural language processing to the work I was doing.
I had a particular focus on artificial intelligence, but I also drafted patents directed to computer hardware, robotics, telecommunications, and even a new cardboard box design.
The benefits of a legal background
In addition to learning about law, law school also teaches many soft skills like negotiation and persuasive writing. In returning to programming, I was surprised to find that the writing skills I’d developed through law school were helpful for writing high quality code. Today’s engineers work in teams, so writing clear and readable code is just as, if not more, important than writing code that compiles.
I think a legal background can be even more valuable to have as a founder for certain kinds of companies. For example, having a lawyer on the founding team of a cryptocurrency company is almost essential, since they must navigate a rapidly evolving policy landscape. Similarly, my legal background is necessary for my current company, which heavily leverages my domain expertise. Because we are building products for other lawyers, my work at various law firms translates into years of valuable customer research.
Today’s engineers work in teams, so writing clear and readable code is just as, if not more, important than writing code that compiles.
Industry Challenges
Law is a collection of very specific industry verticals, but there is a general trend towards using more technology as part of legal practice. About 30% of my work as a patent lawyer was creative, and the other 70% became mechanical over time. A lot of my work required specialized training and practice, but nevertheless became predictable and routine. This phenomenon also applies to other areas of law.
On PatentPal’s goals
The world is innovating at an accelerating pace. This challenges the patent industry to keep up with increasing demands for intellectual property. But patents are time-consuming for humans to produce manually. This was frustrating to me in my own practice where so much of my time was devoted to mechanical drafting and there were very few tools that could help me write more efficiently.
This was likewise a problem felt by my clients. The typical patent takes 40 hours for an attorney to write and costs the client $12,000. This blocks access to the patent system for many individual inventors and startups, who are not able to pay the price of a car to submit a patent application.
We want to solve these problems by building technology that unpack language complexity. We want to simplify lawyers’ workflows so they can focus on what they’re best at. And we want to lower the barrier to entry for inventors so they can viably participate in the IP market.
The typical patent takes 40 hours for an attorney to write and costs the client $12,000.
A common question we get is whether our products will replace lawyers. The short answer is no. Software is very good at automating mechanical and repetitive tasks, and there is a lot of that in legal practice as it exists now. But legal practice is also something that has always evolved. Lawyers who aspire to become the best embodiments of their profession and who seek to deliver client value will naturally hone their craft with the aid of the best tools available.
Lawyers who aspire to become the best embodiments of their profession and who seek to deliver client value will naturally hone their craft with the aid of the best tools available.
What makes PatentPal unique
PatentPal is leading the industry in how much we have automated. Our language-processing engine deconstructs any invention into its basic components — into collections of technical concepts and their relationships. We then assemble these basic building blocks into any document we want to generate. While inventions are often complex, we have also developed tools to elegantly visualize complex arrangements of ideas. Overall, this approach lets users see how our AI has understood their inventions and work with the AI to customize how language is generated. In doing so, users can use our technology to automate workflows while retaining creative control over the generated contents.
Being a solo-founder
Being a solo founder is hard, but also has advantages. This includes being able to deliver a singular, uncompromised product vision and having the freedom to quickly change course in response to new data and changing circumstances. I think the biggest challenge with being a solo founder is managing the emotional turbulence that comes with starting a company. But this was largely solved once I built a team with whom I could exchange ideas. The data also shows that while co-founding teams typically raise more money, solo founders generate more revenue and are more likely to reach a successful exit.
The data also shows that while co-founding teams typically raise more money, solo founders generate more revenue and are more likely to reach a successful exit.
On building a team
We strongly emphasize hiring for culture, especially early on. Something truly magical happens when a team synergizes well. Specifically, we look for individuals who are as excited about the problem as we are, and who bring a ton of energy into the team. We also emphasize being able to communicate ideas clearly and simply, both verbally and in writing, i.e., in how they write and document code.
Future applications of machine learning
Ultimately, we want to build machines that can understand human knowledge. Some of machine learning’s current limitations include its lack of interpretability and its reliance on big data. In other words, we want machines to not only solve a problem, but also explain how and why they solved it. And we also want machines that do not require a billion data points to come up with an accurate solution.
To overcome these limitations, we need to solve the problem of machine abstraction — we want to build machines that understand and work with abstract ideas. This includes creating concepts to describe data clusters, identifying emergent concepts having new properties not found in the sum of its lower-level concepts, and generating names for new concepts that are descriptive of their purpose and contents.
The technology we are building at PatentPal is a step towards this vision. We start with building machines that can understand language, ideas, and inventions within the patent context, before expanding into other domains.
To get updates on this series, please follow our publication on Medium.
Follow us on LinkedIn: https://www.linkedin.com/company/foothillventures
Follow us on Twitter: @FoothillVenture
Foothill Ventures is a $150M seed-stage technology firm. We back technical founders across software, life sciences, and frontier technologies.
Questions, thoughts, reflections? Let us know in the comments below. We’re always looking for great entrepreneurs and early stage ideas, and we’re always interested in having a discussion about venture, technology, and anything related. To see more about Foothill Ventures, please visit our website: foothill.ventures.
