&AI: AI Patent Review

In partnership with

This is The Startup Breakdown, the only newsletter looking into the new Y Combinator companies that you need to know. By joining this growing community of hundreds of founders, investors, and other startup aficionados (think i spelled that right?), you're getting a firsthand look at the companies that may be the next Airbnb, Coinbase, or Reddit.

Intercom for Startups

Join Intercom’s Early Stage Program to receive a 90% discount.

Get a direct line to your customers. Try the only complete AI-first customer service solution.

Apply now

AI Patent Review: The Future of Intellectual Property Analysis

The funniest episode of Silicon Valley has to be The Patent Troll.

The Pied Piper boys are notified that they’re infringing upon a patent, and seeing that the author of the letter is based down the road, Richard decides to go meet who he assumes is a fellow garage hacker to quash the accusation coder to coder.

Instead, he finds an old retired attorney who pays for his lavish lifestyle by doing hardcore patent research and blackmailing founders into paying large-but-not-court-settlement-large sums of money to go away.

For Mr. Burke, spending the 2-4 weeks it typically takes attorneys for patent DD might be a worthwhile investment.

For most lawyers, this time investment, combined with the high chances of missing something in the research process, is inefficient and costly.

 &AI was founded by a pair of MIT alums to address this.

The platform enables instant patent review, pouring over thousands of existing documents and analyzing the current jargon to invalidate or redraft new applications, saving lawyers enough time to come up with new things to bill for to recoup their newfound efficiency.

In seconds, what was once a couple of weeks of work process, is turned into simple, easy to interpret graphs and charts that can guide the argumentative side.

Now, “but there’s so much more to it!” is inadequate as an excuse when your mom tells you that you should’ve been a lawyer for arguing with her.

Market Size:

• TAM: $71 billion

• SAM: $10.7 billion

• SOM: $107 million

Competition:

• Patsnap, LexisNexis PatentSight, IPlytics

• Google Patents, USPTO Patent Database, EPO Espacenet

• ROSS Intelligence, Kira Systems, Luminance

Team:

• Caleb Harris, CEO: Built data science stack @ InsideTracker, CS, Neuroscience, ML research at MIT

• Herbert Turner, CTO: Founding engineer @ Kino AI, LLMs at Google, Aero/Astro, Autonomous Vehicles, and ML research at MIT

Risks:

• Hallucinations: We’re all very familiar with the occasional “put glue in your spaghetti” from AI… even looking specifically at legal cases, there have even been professionals put in time out after using ChatGPT to find legal precedents, only to be provided made up cases…

• Reasoning: Even in a perfect, hallucination-free world, LLMs’ ability to reason, and make judgement decisions that go a step beyond pattern recognition, is limited, and this could cap just how effective outputs can be

• Data security: One of the biggest obstacles to further enterprise adoption of AI has been fears over sensitive information being used with models, and dealing with the private notes and filings of attorneys will make even Harvey Specter sweat

What I like:

• Chat with documents: getting a comprehensive snapshot of all of the necessary info is huge, but being able to dive deeper and ask questions to better understand this data is a whole other level

• Error-specific response generation: Because of its treasure trove of industry data, the tool is able to interpret errors and respond to them immediately and convincingly

• Custom charts: Rather than being another “text in, text out” tool, &AI actually spits out information and data in graphs and tables, making it digestible

Opportunities:

• Case studies: Perhaps the site is clear and makes enough sense to a legal professional, but when I look at it, I get it at a high level, but an example could really sell a potential buyer on the value that the product can provide, ideally for a few different patent cases of different industries (tech, art), or research conducted for different purposes (applying for a new patent, defending one)

• Partner with professional groups: Luckily, if there’s one thing a lawyer likes talking about more than the fact that they’re a lawyer, it’s talking to other layers about the fact that they’re a lawyer… &AI can try to get on the ins with the community, presenting at conferences, getting cozy with the big names, and really building a name within the patent law space

• Partner with other law tech startups: Even within YC, there are at least a dozen other legal tech companies, and leveraging the relationships that these companies already have, whether through content partnerships or through feature integrations, finding a way to provide value to these companies and convince them to introduce their product to their existing audience

How am I doing?

I love hearing from readers, and I’m always looking for feedback. How am I doing with The Startup Breakdown? Is there anything you’d like to see more or less of? Which aspects of the newsletter do you like most?

Cheers to another day,