Latest / Legal Late Night / Revenue Projections for Law Firms & The "10-80-10 Rule" of AI

Transcript

Revenue Projections for Law Firms & The "10-80-10 Rule" of AI · Legal Late Night

• Jared Correia (00:00): Hello everybody. We've got a show that promises to be at least mildly interesting for your listening and watching enjoyment. I'm your host, Jared Correia. I'm the CEO of Red Cave Law Firm Consulting. For my monologue, I'm talking about one of my favorite end of year activities. That's right. Building revenue projections. In the interview, we've got Massyl Mallem of patent plus AI. In the counter program, we're playing the popular family game, Con Man, all about the shittiest of the New England states, Connecticut. Now, let's talk about how to make more money. That's a pretty good teaser, right? (00:45): Okay. As I mentioned, one of my favorite year-end activities that I run through with all my consulting clients is finalizing everybody's Christmas and Hanukkah lists. No, that's not true. The actual year-end activity is creating revenue projections for the firm. If you've never done this before or you've never even heard of this, it'll be a good time to think about doing those kinds of things. Why? Well, because you need goals as a business owner. If you don't have goals, you're just guessing. And assuming is bad. We've all heard the phrase, assuming makes an ass of you and me. So if you're assuming that you're going to make certain amounts of money every year, you're not going to notice trends in your business that could be on a downward spiral. So trend data is something you should review and act on. And one of the ways to figure out how is your business trending both in total revenue and also revenue related to specific practice areas is to actually look at your prior year's numbers and try to project next year's numbers. (01:53): And if this is the first time you're ever doing this, if you just started a law firm this year, then you're doing it for the first time. So I'll tell you how to do it. Now, if you do revenue projections for a while, you're going to have actual results to compare to your projections to see how well you did or how poorly. But that's not entirely what the game is about. The game is largely about setting goals for your business so you can get aggressive and so that your business doesn't stagnate. So let's talk about how you do it. When I talk to people about revenue projections, I basically ask them to reverse engineer a product. So let's take a look at last year's revenue. If you don't have one, you're going to make it up, but we're going to figure out an amount of money you want to make this year. (02:38): And we'll talk about the different types of goals and models in a second. But once you have that number in place, the next thing I would look at is your case values. So you want to know your average case value for each type of case that you have. And then it's just like building blocks. So if you have 20 estate planning cases that you can do this year and you do $5,000 per estate plan, that's $100,000. If you have 10, I don't know, civil litigation cases and you make $10,000 on each civil litigation case, that's $100,000. Stack those two together, you just made $200,000 this year. You can mix and match if you have more practice areas, you can add those in, but you get the general thesis. Come up with a number, try to figure out how that would actually happen in reality. And you can mess around with that in any number of ways you want to. (03:32): You can build different revenue cases as I alluded to before. So you may have an aggressive revenue case, your stretch goal maybe. You're like, "Boy, I would really like to make this. " And then run the numbers and see, well, what would that look like? What would I have to do to hit that number? How much more money would I have to make this year than last year? You could run a breakeven case like, okay, what do I need to do? Let's say it's your first year running a law firm or it's your first year hiring associate. How do I break even? What's that number? You could also, if you're an attorney, which you probably are, if you're listening, and looking at sort of like the worst case scenario, which is whatever your attorney does, you could even figure out how you manage for that. (04:13): What if I make half of the money I made last year? What would I do in that situations? How could I react? So if you start looking at this and using your average case values as a building block, and you can probably find that number in a case management software and accounting software. So this is not all hand work. That should work for billable hour firms and also project based firms or flat fee firms. Now, just remember, the idea is to not guess with perfect accuracy. So you're not trying to hit a number necessarily, a specific number. You're trying to make an educated guess and ideally you have a stretch goal in place that you're trying to hit to make yourself work a little bit harder at this. Probably harder than you've worked if you've never done revenue projections and if you're focused on business growth. (05:05): If you do this quite a bit, you start doing it year over year, you'll probably get closer. Every year you do it. So again, it's not perfect. It's just an idea of what you'll make. If you've done this previously, what's cool is that you can compare your projected revenue to your actual revenue. How close was I? Was my guess insane? Was it fairly accurate? Maybe I didn't stretch enough on my stretch goal last year, so I need to be more aggressive about that this year. I know when I talk to clients of mine, I have very different ideas about what their stretch goal is versus what they think their stretch goal is. Mine's always higher. Now, if you have a contingent model, this might be a little bit different. So for those, the payoff is longer term. You're funding those cases. So there's some additional costs, but in terms of like top line revenue, like how much money you can make, the criteria you should be looking at is the case settlement value. (06:07): And again, this is just educated guesses, projections. What do I think this case is worth? And then you also want to have a projected settlement date. And then what I like to do with contingent firms is I also like to put a potential for settlement number on there. If a case is 70% likely to settle versus 30%. And just as a backstop, I have a potential trial date and a potential trial award just in case the settlement doesn't happen. And obviously when you're looking at these numbers, you can see, okay, what's the likelihood that this is going to trial based on my percentage to settle figure? So that's a good starting point. And you can build it out from there and add more data points if you want to. But the difference is you need to extend it out over the course of years because some of those contingent cases take two, three years to settle. (06:56): So I would have, at a very baseline level, spreadsheet, workbooks for different years, and then I would spread those numbers out so you have a good idea of what your cash flow looks to be over a two, three year period. So slight tweak on that. And again, you can have stretch goals, you can compare that to actual, a whole host of things you can do there. But if you don't have a revenue projection, you don't really have a goal as a baseline goal for the year for your law firm. And I think a lot of goals can cascade off of that. So as I just alluded to, I think the real value of revenue projections is the conversations they force you to have. So you have to figure out, if I need to get 10 more cases, how do I do that? Well, that's a marketing question. (07:46): So you have to figure that out and you probably have to talk to your team about the strategy. Or if you're a solo attorney, you have to meet with yourself. How many leads are you going to convert and from what sources? Okay, now that's an intake and conversion conversation. So who do you talk to about that? What kind of software do you use? Do your pipelines as you've set them up work effectively? What are those baseline data sets to begin with? Like what's your lead conversion rate right now? Who's going to bring the work in? Is that going to be built on paid marketing and advertising? Is that going to be an origination thing where you're relying on attorneys to bring work in using referrals, the traditional law firm strategy? And then once you get the work, who's going to do it? So now we're talking about work allocation, case allocation. (08:31): Who do I give this to? And what are the criteria for passing work into the firm and can people share origination fees? What's your profit expected to be? Now we're talking about a different thing. We're no longer talking about top line revenue. We're talking about how much money you're going to actually make after taxes, after your budget. So you have to figure for all that stuff. And oh, by the way, if you don't have a budget, this is a good time to run a proposed budget as well so you can get a profitability number. And then you can compare that to the actual budget for the year in the same way that you're comparing your revenue guesses to the actual revenue for the year, et cetera, et cetera. Like everything else, data doesn't mean shit unless it generates conversations that lead to action. Speaking of conversation that leads to action, we've got Macil Malim of patent plus AI coming up right now. (09:26): We're talking about patent searches and we're also discussing balene whales. Weight being told only one of those is true. Although I suspect we'll talk about the Hartford whalers, be back Jared Correia (09:36): In a second. Well, I've effectively run out of things to say, which is awkward Jared Correia (09:46): Because this is a podcast. So I'm just going to go out and run some errands. I'll be right back. Nah, that's kind of weird. Let's bring in our guest instead. I can always pick up some milk later. Our guest today is Massyl Mallem, who is the CEO and co-founder at Patent Plus AI. Massyl, welcome to the show, man. How are you doing? Massyl Mallem (10:10): I'm doing good. Thanks for having me, Jared. I really appreciate it. And if you want to go run some errands, I'll take over from here. It's okay. Jared Correia (10:15): Oh great. Okay. I'll see you later. No problem. You taught me a little bit of Arabic before we started, so I appreciate that. Massyl Mallem (10:22): Of course. Anytime. I actually have Arabic Twitter on the side to help fund this startup because startups don't make money. Jared Correia (10:28): Oh, yes. We'll get into that. I thought I was a little bit more badass with the Arabic than I actually turned out to be. So my bad Massyl Mallem (10:36): On that. I'm still earning. Yeah. It's just English is very limited with the kind of noises that you can make in the English alphabet. So there is so little you can do to undo years and decades of pronunciation, but there's at least like six different sounds that exist in Arabic alphabet that just don't exist in romantic languages. Jared Correia (10:58): That's wild. My son is a teenager, so he makes a lot of sounds, mostly different variations on grunting. So let's talk about the startup thing. Why on earth? Would you jump into a startup company? It's not because of the money, right? Massyl Mallem (11:18): No. Yeah, not that. No. Actually, I do entrepreneurship teaching at Yukon every once in a while, kind of just talking about my experience and whatnot. And I always, when I get the opportunity, I tell students not to pursue entrepreneurship. But with the caveat of, if you want to do it because you want to get rich quick, then it's the worst decision. But if you want to do it to learn, then it is the best and there's nothing that can teach you as much as entrepreneurship. Both in terms of the business, but also a lot of the technical world. There's very little you can do to learn quicker than just by applying it, doing it yourself and learning as you go. So intellectual curiosity is definitely a prerequisite to any entrepreneurial endeavor. And a lot of times, money does eventually come across and you get lucky, but you need a lot of luck as well as dedication and smarts and everything. Jared Correia (12:14): I think that's a cool way to look at it from an intellectual curiosity perspective. I don't hear people say that a lot. I hear people mostly be like, "I'm going to sell out or I'm going to make a ton of money doing this. " That's a great viewpoint that you have. Okay. So why did you decide to do this? Because I don't think you practiced law. I think you have an engineering background. Correct me if I'm wrong. Massyl Mallem (12:38): Yes, sir. Jared Correia (12:39): How do you go from engineering to legal? Massyl Mallem (12:42): Yeah. Well, intellectual curiosity. Actually, in high school, I love political science and history, but when it was time to pick a college major, I was like, right, I need to work. So I went more technical route. I studied chemical engineering. And from intellectual curiosity perspective, I was interested in so many majors. It was very hard to pick, but eventually I picked chemical engineering and then I saw an ad, was actually studying for organic chemistry and obviously got tired, decided to procrastinate a little bit. I opened up my email, I see a job posting for intellectual property. And in the job posting, they're like, "We're looking for either a PhD in chemistry or a pre-law student." And I'm like, "Why would you put pre-law and chemistry in the same sentence?" And I always wanted to be a lawyer since I was little. And obviously I was studying in chemical engineering. (13:31): So I was like, "This sounds perfect. It's like the marriage of the two things that I like the most." Again, from an intellectual perspective. So decided to take the opportunity, interview, and it was like a full-time position as an intellectual property specialist. The person who had the job before me was going on maternity leave, and I guess they were like, "Why not? Let's replace her with a sophomore in college and see what happens." And it was honestly just love at first sight. I kind of found my call in to be a little cliche and I was doing everything from working with inventors on invention disclosures, all the way to working with attorneys on opposition and invalidation cases. So I got really kind of like a holistic view on the entire intellectual property world specifically for patents. And I got really focused on patent searching. That was kind of the most repetitive task that I was doing and I wasn't satisfied with the options available to the industry in terms of solutions for patent searching. (14:29): So my best friend and I, or college roommates at the time, decided to give the industry a better option to perform patent searches, and hence we founded the startup. Jared Correia (14:37): Okay. Let me tell you a quick story about myself. I took organic chemistry in high school as a junior. Massyl Mallem (14:45): My condolences. Jared Correia (14:46): My grade was 35 at the end of the day. And the final was a video game. 35 out of a hundred was my final Massyl Mallem (14:54): Grade. Jared Correia (14:55): I know jack shit about chemistry. So I'm impressed already that you can even do this shit. Okay. So can we step back one second before we get into the software? Why did you decide to get into chemistry? That seems like a terrible choice. Massyl Mallem (15:13): Yeah, no. So again, in high school, I hated chemistry. Absolutely despised they burn in passion. And I had a run-in joke. I was like, why would I need to know the charge of boron? And I'm like, I would never need to use that. I want to be crazy. Jared Correia (15:29): I asked myself that question every day. Massyl Mallem (15:32): Right? Yeah. So for me, it was completely useless. And then I remember for the senior year, I had really, really bad grades and I had to get like a hundred or a 95 or whatever, like an A plus on the final to pass the class. So I'm like, I guess I have to learn this now. So I just applied myself and just had a whole binge study session to kind of get my knowledge set. And it was just general chemistry, nothing crazy. And I liked it. The more I learned, it's like a game. It's like learning a language essentially. And for me, I've went from ... My first language was French and then my first language became Arabic and now my first language is English. So I love learning languages and chemistry was essentially that, kind of just like a game, a lot of rules and it felt like learning a language. (16:19): And I enjoyed the process of studying for that final because I needed to get a perfect score. I ended up getting a perfect score and I passed the class and everything. So when it was time to pick a college major, I was like, well, I have mechanical engineering, chemical engineering, electrical engineering. Electrical seemed a bit intimidating and chemical just had better job prospects. At least as a 17 year old, those are kinds of the conclusions that I made. So I narrowed down the top three that I will do is electrical, chemical or software. Software sounded boring. I'm like, what? I'm not going to stand or sit in front of a computer screen and write code all day. That sounds awful. I would never do software engineering. And so I was like, I guess chemistry, I have this little love, hate relationship with it, so let's give it a shot and see what happens. (17:01): In retrospect, I'd say I narrowed it down to the best three. I would agree with that, but probably electrical and software are a little bit more valuable degrees than chemical engineering. But for me, it was a practical decision. It was like, what is the best return on investment from a bachelor's degree? And those are the three that I narrowed that down to. Jared Correia (17:19): I like how you were like, "I would do software engineering, but it would be more exciting to potentially electrocute myself." Massyl Mallem (17:26): Right, literally. And then as things happen, it's like you're making your own plans and God is making plans completely different. And usually whatever you're saying, "I'm never going to do this, that's exactly what you're going to end up doing." Same thing with startups. After I did engineering, I was studying full-time in school of engineering, I was like, plan A is like a patent agent, plan B, patent agent to work for a law firm because I wanted to go to work as a patent attorney after that. Plan C was work as an engineer, plan, whatever. Startup was kind of just like a fun thing on the side. I never planned for it to take off. It was kind of like, this is my way to stay in touch with the IP space because I loved it so much. But I was realistic and I'm like, 90% of startups fail. (18:09): So it was like plan D. But eventually we kept doing more competitions, kept getting more grants. Eventually we got VC funding. So I was like, I guess we're doing it full-time. Let's see what happens. And now it's like I can't imagine doing anything else. Jared Correia (18:24): Yeah. Now it seems to be a good time to mention that Einstein was a patent clerk, as I'm sure you know. Massyl Mallem (18:29): I was not aware of that. Jared Correia (18:32): So let's talk about this. It seems like you do do a lot of competitions. Has that been a viable way to get notoriety in the space? Massyl Mallem (18:43): I would say, yeah. Yeah. Actually, yeah, well, different types of competitions. So we started off originally doing a lot of entrepreneurship, like pitch case competitions. They have a lot of those in university and they give you a problem, come up with a solution, you make a pitch. We did that a bunch of times, made a little money off of that. It was just like freshman year, we're still 18 years old and just the process of ideating, of coming up with solutions, developing MVPs, developing prototypes, that was kind of very fulfilling and that was kind of our intro to entrepreneurship. More recently though, we've started doing a lot more serious competitions, which if you're familiar with unified patents. So they're basically like an organization, all different types of Fortune 500 companies essentially pay them like a fee on annual basis. And whenever one of these companies gets challenged by a non-practicing entity better known as a patent troll, troll patent, they go to unified and they unify takes the patent trolls patent number and they put a bounty on it essentially. (19:42): So they're like, anybody in the world, if you can come up with prior art that can invalidate this troll patent, we'll give you 5,000 or $10,000 or whatever it is. It sounds Jared Correia (19:50): Fucking great. Okay, go on. Massyl Mallem (19:52): Yeah. Jared Correia (19:53): Hunting patents. All right, Massyl Mallem (19:55): I'm on that. And non-practicing entities for the non-initiated are basically just organizations that pursue intellectual property rights for the sole purpose of suing people. So they're not interested in innovating or building or selling, it's just so I can attack companies and whatnot. And they're a big problem, especially in the US. So there is a lot of incentive for people to do this. And for us, the system that we've built is, and I should just backtrack a little bit. So patent search in general, you have two types of solutions, and we can dig a little bit deeper into this, but you essentially either outsource your project to a search firm and let them handle it for you, or you purchase subscription to search software and you do the search in- house. You being an IP associate at a law firm or a company or whatever. Paralegal Jared Correia (20:41): Associate attorneys, right? Massyl Mallem (20:42): Yeah. Yes. Usually people need to have some sort of technical background to perform the patent search. We're the first company in the world to give the industry a third option, which is don't outsource, don't do it yourself, instead automate, automate with artificial intelligence. So our solution is not a search engine that people use to search. It's an autonomous search system that performs a search on behalf of the user. And for us, yes, it's cheaper and faster, but those are just side effects. For us, this is a quality play. For us, humans are very bad at patent searching, AI systems are significantly better. So we want to give folks and clients a higher quality patent search, which is why we developed the system. So to tie it back to unified patents, basically we take these patent tools, numbers, we put them through the system, get the claim, invalidate it, submit the prior R2 Unified. (21:33): It's intelecommunications the last one that we won. And I don't know anything about telecommunications. So the system completely just understands the patent, invalidates it. We submit the prior art and we end up, we won multiple of these competitions at this stage. And yes, it has been a great way to get A, notoriety. A lot of clients kind of reached out to us after they heard about that and they're like, "Wow, this sounds great. Can we use this system ourselves?" But also to prove to folks that the quality is good because we can do a lot of internal benchmarks and internal testing, but those are whatever, especially a lot of the benchmarks, the back tests that we do are against the USPTO or the EPO patent databases or like, "Hey, the examiner thinks this is the best prior art." Our system also found that, but fundamentally examiners are really bad at patent searching. (22:19): They're by no means the gold standard. Why? Because they're humans and humans are bad patent searching. So we're like, no. And for example, a lot of, especially recently, patents that have been granted by examiners end up getting invalidated in post-trial proceedings because inventory-throwing Jared Correia (22:34): Everyone under the bus today. It's a large bus on the podcast for everyone. Massyl Mallem (22:39): Yeah, exactly. It's double decker. Jared Correia (22:41): Everyone sucks. Massyl Mallem (22:44): Except, well, even AI search engines suck if not used properly. But yeah, I mean, for now it's really just one company that's offering autonomous patent searching. So you've Jared Correia (22:59): Got patent plus AI, but then you also have ... Part of the feature of that is like Autopad, which is essentially your AI agent. So if somebody's out there listening to this, they could think of like Lexus and one of their co-counsel research tools, right? Massyl Mallem (23:18): Essentially. Well, Autopad is really the bread and butter. The Patent Plus AI is the company name, but Autopad is the star of the show for sure. And that's the name of our system that performs the searching. Exactly. And to wrap up on that, so basically we go to folks, "Hey, you can take our word for it or you can see these multiple troll patents that we've involved, the Autopad invalidated." I'm glad you brought that up so now I can refer to it, but it's actual name, AutoPad validated. So okay, I can see it. It's not just the internal tests or internal benchmark, it's a public. And the thing with the Unified is you'd have dozens of search firms try to invalidate these stroll patents. A lot of them also from India, because obviously 5,000 or $10,000 goes a long way there. So it's expert search firms, very technical, technically capable, and they try and they fail, but Autopat finds better prior art because it's not a human doing the search. (24:13): And we can dig a little bit deeper into the details of why and how. How much do you Jared Correia (24:18): Hate humans? No, I'm just kidding. Massyl Mallem (24:22): AI never sleeps, Jared Correia (24:23): Doesn't need Massyl Mallem (24:24): Snacks. Yeah. Well, no, it's actually out of love for humans that we've developed the system and we're automating this part of the job because the way we talk about automation is this thing we called the 1080, 10 rule. So these are percentages. We believe every patent search project and all other types, most of intellectual projects can be split in discount, can be understood in this kind of framework. So the first 10% and the last 10% is where our clients get involved. The middle 80% is automated by Autopath. So what that means essentially is the clients get involved, say for example, for invalidity, they'll get involved in the first 10%. This is the competitor claim that I want to invalidate. These are the most important elements. These are the features, these are the must haves, nice to haves, et cetera. If there is any ambiguity and patennlized by definition ambiguous, they can give custom instructions to alleviate the ambiguity, tell the system how to interpret things, basically set up the search strategy. (25:27): And then Autopet comes, does 80% of the work, the bulk work, the tedious work, and then the human expert comes back at the end, the final 10% to make IP decisions. Do I file, do I not file? Do I invalidate? Do I not invalidate? Do I do a follow-up, such, et cetera? Why? Because that's what humans are good at. Humans are good at abstract thinking, strategic reasoning and decision making, i.e., The beginning and the end of the process. The middle, having to read hundreds of patents to find five relevant ones, having to try dozens, if not more, search strategies to unearth the best prior art. Humans are bad at processing that kind of data, especially when each patent is like a 50 page long document filled with dense technical literature and legal jargon, sometimes on purpose, because patent lawyers, lawyers in general are sociopaths, as we all know, and they like to complicate things more than necessary. (26:23): I know, probably a very popular joke with the listeners of this podcast. Jared Correia (26:28): I don't know that I've ever heard anyone call lawyers sociopaths from the show before, but I'm down. Massyl Mallem (26:33): Let's explore it. Why not? Yeah. Well, I mean, because what I mean is obviously sociopaths are smart, so it's a compliment, so to speak. Please be my clients. I love you guys. But the idea is sometimes the language is overly complex on purpose to make the patents harder to find doing a patent search, and then they come back and they can sue you for rights, et cetera. So the point is, it's complicated from a technical perspective and the legal perspective. And then there is thousands and millions of them that you have to review. Humans can't do that, but humans can strategize and make decisions. What AI can do is process that kind of data, sort through the legal jargon, sort through the technical jargon, and do that ad nauseum without taking a break, without picking up its kids from daycare or whatever it is that humans do, which is why we make it dictated towards the 80%. Jared Correia (27:29): I feel like that's a shot at me, but I'll move past it. So let's talk more about the sociopaths. Let me just speak what a lot of lawyers are probably thinking right now as they're listening to this. Sounds cool. How do we avoid hallucinations? Because that's a big concern of lawyers when it comes to AI. Massyl Mallem (27:51): Yeah. Well, for us, that was a problem maybe back in 2021, in 2022, but right now this system is literally incapable of hallucinating because it doesn't- Oh, incapable? Incapable. Incapable. It can't, because it doesn't generate text, it doesn't generate citations, it doesn't generate quotes, it just cites them. So for example, for a particular invalidity, for feature that's anticipated, the system will say, "Hey, it's anticipated by this quote, this quote, this quote." The quotes are verbatim cited from the patent. They're not generated by the system. And for people who use our web app directly, because we have different business models, some people just work with us as a search firm through email, but most clients use the web directly, they actually just click on the quotes that are cited and it will take them exactly to where in the full patent description those quote quotes live. (28:44): So they can literally see exactly where each quote comes from. Nothing is made up from thin air. Jared Correia (28:50): And I forget, I'm just asking because I haven't been ... I don't think I've ever worked in the patent space before, that my wife works at the patent law firm. Do you have to pay for USPTO searches? Massyl Mallem (29:03): No. Well, funnily enough, actually, the USPTO, as of a month ago, launched a new service of automated AI searches. It's garbage so far. Everybody hates it. But yeah, so generally you're not required per se, but you're very, very encouraged to do a search before you file. Oh, right, Jared Correia (29:23): Right, right. Yes. Massyl Mallem (29:25): Exactly. In the private space, go find someone or do it yourself using whatever search engine to just make sure that you're not spamming them with the garbage applications. And then the examiners will have to do the searches themselves. But right now, because everybody's spamming them with applications, especially with people using AI, not only to draft patents, but also to invent, the rate of applications is going through the roof. Jared Correia (29:49): We're already hitting vibe patents, huh, already? Massyl Mallem (29:53): No, literally, exactly. Yeah. It's bingo and you have vibed patent drafts. So the office is completely overwhelmed with how many applications they have. And honestly, for us, this system originally was developed with governments in mind. For us, back in 2019, when we first came up with the idea, the problem that we really wanted to address is that examiners are completely overworked. The average patent application spends two weeks on an examiner's desk when it's actually prosecuted, they're doing the search, making the decisions, et cetera. But it takes two years or more for the applicant to hear back on the AFIS action, to hear back from the examiner. Why? Because there is tens of thousands of application backlog at the examiner's offices and it's only getting worse because like I said, people are using AI to draft and people are using AI to invent even all types of. Jared Correia (30:50): People are assholes. Please go on. Massyl Mallem (30:52): Right. So the backlog is only and only getting worse. So for us, it's not people are like, "Oh, how could you automate this search of the PTO?" For us is how could they continue to function without automation? Specifically, like I said earlier, the 10, 80, 10 rule, because we're not going to have AI tell people you get a patent, you don't get a patent, that's a recipe for disaster. Instead, the examiners are involved in the first 10% to set the strategy. AI gives them the best references that it finds, and the examiner comes back at the end to make that decision of whether to grant or to reject the patent application. And that's what they're good at. Examiners are not good at patent searching, but they're good at decision making. They're good at telling you, "Hey, this is gets it doesn't..." But actually finding the prior art, they're very limited because of their human brain. (31:39): Actually, they did a study, I think, a couple of years ago where they gave three examiners the same pending application, the same pending claims, and each examiner, when they search, they came up with different prior art references and different rejection arguments for that. Why? Because they have biases. Biases based on their education, biases based on the last case that they did. Sometimes they get lazy, this is good enough. That's not how our system works. It doesn't have biases. Instead, it will let the data, the patent database, influence decision making. It doesn't go into the patent database with already some preexisting biases. It will search, add nauseum till it finds the objectively best prior art, not just go off of baseline experience. So that's why for us, this tool is perfectly utilized by patent offices from across the world. Jared Correia (32:32): Where do you go from here? I think this could take a lot of different directions. What next feature sets are you working on? I mean, I suppose I could see a consumer facing tool. So what's next for the software? Massyl Mallem (32:46): Yeah, we do work with the entrepreneurs and startups for a D2C route. It's nice to fulfill them, but honestly, it's just startups will do maybe one, two, three applications in a year. That's like three patent searches. And for us, the system is built for volume. It's built for doing as many searches as possible. So it's just for us not too much return on investment working with startups. In terms of finding them, once they have the projects very easy, because they do very simple invalidity and novelty searches, which are the easiest to do. We do them in 12 hours. When I say we, I mean Autopat have been searched in years, but the system is more so built for volume. In terms of what next, for us, we've been asked to do drafting, we've been asked to do a bunch, and we will never do that because we are laser focused and locked in on being the best in the world at searching. (33:37): Everything else is a noise. It's a distraction. Other people are innovative in it. For us, we found our niche within patent law and that's what we're focused on. And we started off just with patentability and novelty searches, and then we moved on to doing invalidity searches. And now the system can also do freedom to operate or clearance searches. And all those with different degrees of autonomy and just making this system better and better. And being able to do these at scale is life changing to a lot of clients. Especially, for example, for FTOs, a lot of times clients are making the budgets, be the primary decision maker on whether they do a clearance search or not. And as you likely know, with FTO searches, it's basically a risk factor. It's like, how risky do I want to be? Do I just go ahead and start selling or making this product without making sure I have the clearance to do that? (34:30): Well, it costs 10,000 to do it, so I'm probably fine. They go ahead, launch the product, then they get sued for millions of dollars. So being able to offer folks unlimited FTOs at scale at a very reasonable cost, that's going to help them mitigate some of that risk factor in their IP decision making. And that's the goal at the end of the day. Jared Correia (34:54): In two minutes or less, will there be AI lawyers at some point in the future? Massyl Mallem (35:02): I will have ... Yeah. Well, I don't think there should be, but it's, again, to revert back to the 10, 80, 10 rule, I think that is unquestionably inevitable. You think Jared Correia (35:13): Always human in the loop? Massyl Mallem (35:15): Yeah, no, for sure. Well, and then the loop is specifically the beginning and the end. And I think that's not just for lawyers, for almost all intellectual professions. It's just humans are good at strategic thinking, abstract thinking, and decision making. That's where their expertise should be isolated, and that's where their human factor can best be put to use. So again, it's not automation because we hate humans, it's automation because we love humans and we want to help them focus on what they do best. For example, a lot of search firms are our biggest customer segment. They're not the biggest, probably the second or the third, but we love working with search firms. At first, they hate us like, "Oh my God, you're coming from my job. You're stealing my job." Whatever. We're like, "No, no, no. We're making your job fun." Because when you ask the search firms and the patent searchers what they like most about patent searching, it is the setting the strategy and making the decisions. (36:07): Those are the fun parts. Writing endless queries or reading 100 irrelevant patents, that's not what they like about their job. Same thing with lawyers. There's a lot of monotonous, a lot of tedious middle ground that they don't like doing. That's what AI is best primed to do. So I think a version or another of the 10, 80, 10 rule is probably what the future looks like. Jared Correia (36:31): Last question for you. How's the Brazilian jiu-jitsu coming along for you these days? And could you kick my ass if you had to? Massyl Mallem (36:40): No, I wouldn't. I wouldn't, but it's been ... Honestly, man, I wish I had more time to do more BJJ. I had time in college, but it's just entrepreneurship is ridiculous. I start with a day with 14 items on my to- do list, and then after 10 hours of working, it's 16. You know what I mean? So it's been very, very busy. Especially now with snowboarding season coming on, it's like jiu-jitsu is going to take a backseat, for sure. But it's definitely going to always have a special place in my heart. Jiu-Jitsu and chess are like, I think the two sports that lawyers enjoy the most because it's the same kind of principle of anticipating moves. Jiu-Jitsu is a chess game and chess is a jiu-jitsu game and law is a combination of both of those. So yeah, always one of the special places in my heart. Jared Correia (37:30): Ms. Seal, thank you. That was fun. Can you hang around for one last segment? Do you mind? Massyl Mallem (37:36): Absolutely. My Jared Correia (37:37): Pleasure. Jared Correia (37:38): All right. We'll come right back. All right, everybody. Welcome back. It's The Counter Jared Correia (37:47): Program. It's a podcast within a podcast. This is a conversational space where we can address usually unrelated topics that I want to explore at a greater depth with my guests. Expect no rhyme and very little reason. Massiel, welcome back. Massyl Mallem (38:03): Good to be back. Thanks for having me, Jared. Jared Correia (38:06): I don't know if we covered this in the first half or if we covered this pre-show, but you're a Connecticut dude. Yes, sir. Even though it's the shittiest state in New England. That's Massyl Mallem (38:16): Right. Yeah, but it's the basketball capital of the world. Jared Correia (38:21): All right. Let's come back to that. I want to play a little game I've invented with you. We're going to call it Con Man. All right. Con with two Ns, just so everybody's aware. Basically, I'm just going to ask you Connecticut questions and you can wax poetic on those. Okay. Massyl Mallem (38:36): Are you prepared? So I'm born and raised in North Africa. So that's my original alba mater, but I've been a Connecticut resident for the better part of a decade, so I think I should be okay. Jared Correia (38:48): Yeah. So let's test your bonafides. Okay. Massyl Mallem (38:50): All Jared Correia (38:51): Right. I'm a New Englander as well, as you know. Although Connecticut is kind of like New York to me. What do you think? Do Connecticut people feel the Massyl Mallem (39:00): Same way? Did you know it's called Connecticut because it connects between Boston and New York? I actually did not know that. No, that's a lie. Oh, shit. I was like, "That's great." Right? It Jared Correia (39:14): Makes so much Massyl Mallem (39:14): Sense. Jared Correia (39:16): I'm like, "I got to go find a YouTube video on this. " I kid make fun of me because I watch videos on YouTube about geography. They're like, "You're such a fucking loser." Massyl Mallem (39:25): I love that. I love that. I do that too. Big, big fan of journey. We actually stole the name from some Native American thing. That's where Connecticut comes from. Yeah, Jared Correia (39:35): It's a very New England thing. Literally. All right. Let me ask you my first question. What is the appropriate name for a submarine sandwich in Connecticut? Massyl Mallem (39:50): Subway. Subway was invented in Bridgeport, right? Jared Correia (39:53): Yes, I believe that's correct. So what do locals call it? Do they call it Hoagie, Grindr, Sub? Massyl Mallem (40:00): Grindr. Grinder or Sub. All right. Jared Correia (40:03): All right. You got to ... Massyl Mallem (40:05): Let's go. Jared Correia (40:05): Do you know why it's called a Grindr? Massyl Mallem (40:07): Absolutely no idea. Jared Correia (40:09): Okay. Now I feel like I should just make something up to fuck with you because you made something up about the kit, but I'm not going to do that for you because I'm a nice guy. Or maybe you couldn't Massyl Mallem (40:20): Think of something. Jared Correia (40:22): So there are a couple different theories. The first theory is that the bread is hard to chew because it's like crusty Italian bread, so you have to grind it with your tea to eat it. And then the other theory is that Italian-American dock workers were known as grinders and they named them after the sandwich. Massyl Mallem (40:40): Interesting. Interesting. Never heard that. I think it's probably the former because I don't think Subway was invented by Italians, was it? Are there even Italians in Bridgeport? Were there Italians in Bridgeport? Jared Correia (40:53): I don't know. We'll have to find out. Well, I'll get on that. I'll follow up with the next Massyl Mallem (40:57): Step. Get on the YouTube video and let Jared Correia (40:58): Me know. Yeah. No, I fucking won't do that. Don't expect that, anyone. Okay. Question number two. What is the term that is used for a liquor store in Connecticut? What is shorthand for a liquor store? Massyl Mallem (41:12): I don't drink alcohol, so I don't- Good. I can help you. Many liquor stores, right? Jared Correia (41:18): Have you heard the terminology package store or- Massyl Mallem (41:22): No, I Jared Correia (41:22): Have not. Oh, really? Okay, good. I can teach something. Massyl Mallem (41:24): Wait, Packer rings a bell. Packet rings a bell. Jared Correia (41:26): Yeah. I don't want you walking around Connecticut sounding like you're not local. So the package store is a very New England thing. It's up here too in Massachusetts. And basically it stems from blue laws. So back in the day, people had to conceal the alcohol when they brought it out of the store. So it had to be in a brown paper bag. So that was called the package and that's why they're called package stores. We're Massyl Mallem (41:51): Doing well so far, I think. Makes sense. Jared Correia (41:52): How you feeling? I got a few more for you. Massyl Mallem (41:54): Sure. Okay. Let's do it. Jared Correia (41:57): I'm ready. Next question. Preferred route to either Massachusetts or New York. I- 95 or the Merritt Parkway? Massyl Mallem (42:12): The answer is I just put it on Google Patents because I'm a Gen Zer. Okay. But, but, but I will say- Come on, man. Jared Correia (42:18): Fucking killing me. Massyl Mallem (42:22): I will say I- 95 is probably the correct answer. Oh my God. Jared Correia (42:26): Really? Massyl Mallem (42:28): Yeah. It's the two lane or the three lane highway, but it doesn't go all the way up Jared Correia (42:34): To- Two lane highway. Boston. It's got scenic overpasses. That's the thing. Scenic overpasses. So do this for me. Take a quick trip on the Merritt Parkway. Like you can see every overpass is different. It's designed slightly differently. It's scenic overpasses. And that's the shortcut I always take to go to New York. Massyl Mallem (42:56): No, Jared Correia (42:56): It does not go all the way to Boston, but it's a portion of Connecticut. Massyl Mallem (43:01): Got it. Got it. Jared Correia (43:01): Oh, we got Brown today. This is exciting. Massyl Mallem (43:04): Yeah. I'm just L after L after L. Let's see if I can redeem myself with this next Jared Correia (43:08): Question. Next question. Who has the best college basketball team? Yucon dude. Massyl Mallem (43:16): Right. Jared Correia (43:18): As I look at the giant banner behind you. Massyl Mallem (43:22): Yeah. We actually did undergrad research in Duke. I lived there for four months in Durham. Jared Correia (43:28): I know. That's why I wanted to ask you. You got to sell out one city. Which one are you going to sell out? Sounds like Durham. Massyl Mallem (43:35): I will say maybe Duke, maybe last century, maybe the early 1900s, that's the correct answer, but in the 21st century, the question is obviously Yukon. Jared Correia (43:46): That is fucking ice cold. Okay. Let's stick with sports. What is the name of the former professional hockey team that used to play in Connecticut? Massyl Mallem (43:58): You're killing me. I'm a football fan, the original football, the European football. Jared Correia (44:05): Stones left and right. Massyl Mallem (44:08): The only American sport I Jared Correia (44:10): Watched. I will give you multiple choice. You can guess, but go ahead. What's the Massyl Mallem (44:15): Only American Jared Correia (44:15): Sport you watch? Basketball? Massyl Mallem (44:17): College. College basketball, not any basketball. I do watch MMA though. Big fan of MMA. Who's your Jared Correia (44:24): Favorite MMA fighter? Massyl Mallem (44:26): Well, Habib, before Jabim Normago Marov, before he retired, now his cousin, Islam Makachev is reigning champion. He moved up a weight class and now he's double champion. So he's making all of us proud in the Muslim community. But yeah, outside of MMA and college basketball, I'm a big Chelsea fan. Again, that's in London, European football Jared Correia (44:46): Or soccer. I never watched a single minute of MMA, believe it or not. Massyl Mallem (44:50): You're missing ... Oh, MMA. Jared Correia (44:52): Got you. All right. Massyl Mallem (44:55): As American, as they come, I was going to say, because UFC obviously put MMA on the spot and Dana White literally just built it up from the very ground up. So very, very American sport. But I will say ... Go ahead. Jared Correia (45:09): So you just let me know when there's a good MMA fight and I'll check it out. Massyl Mallem (45:13): Okay. No. Jared Correia (45:14): Do you want multiple choice for hockey team? You might be able Massyl Mallem (45:16): To guess. Yeah. I'm good at multiple treat questions. That's how I passed. Engineering Jared Correia (45:21): Schools. Let's test that theory. So is the former professional hockey team from Connecticut called the Berlin Blazers, the Stanford Stampeders, the Hartford Whalers or the Mystic Pizza? Blazers, Stampedters, Whalers. Pizza. Massyl Mallem (45:42): Blazers. Yeah. I think ... I don't think it's the Stanford one. I feel like Stanford is such a new-ish city. I don't think they've had a good hockey team and adopt. It's just you're too close to New York. I don't think it's Stanford. I don't think it's Hartford. I think it's Berlin Blazers. I think it's Berlin Blazers. Jared Correia (46:00): I was Massyl Mallem (46:00): Proud of Jared Correia (46:00): Myself. I came up with Berlin Blazers all on my own. Massyl Mallem (46:03): Oh wow. It Jared Correia (46:04): Sounds like a team, right? Hartford Wales. Massyl Mallem (46:07): Yeah. Hartford Whalers, yeah. I guess I should have gone with Jared Correia (46:09): Hartford. Hartford Whalers has one of the best sports logos of all time. Should check it out. Interesting. But they left Connecticut a while, probably like 25 years ago at this point. Massyl Mallem (46:22): Yikes. Jared Correia (46:22): And they had a band. They had a brass band that would play a song called Brass Bonanza. That was their theme song whenever they scored a goal. Massyl Mallem (46:32): Interesting. I was going to say 25 years ago I was barely ... I couldn't even talk yet. So yeah, I didn't get a chance to. I'm Jared Correia (46:42): Going to go walk into the woods and die now. Massyl Mallem (46:47): Let's do one more. I Jared Correia (46:48): Will Massyl Mallem (46:49): Say Jared. World Cup is going to be hosted in America this summer. So if there's any time Jared Correia (46:54): To ... Has some games, Massyl Mallem (46:55): Right? They do, but a lot of the semifinal and the finals are going to be in New York. You're going? Or New Jersey, definitely. I will try to. A little bit expensive to find the tickets and everything, but I'll definitely try. Algeria is playing. So I'll definitely try to catch a game. Jared Correia (47:11): I like how you're trying to redirect this to stuff you care about, but this is my Massyl Mallem (47:14): Show. Yeah, exactly. All Jared Correia (47:16): Right. I got one more for you. This is probably your favorite TV show of all time. So let me ask you, where did the Gilmore Girls live in Connecticut? Massyl Mallem (47:26): Do Jared Correia (47:26): You know what the Gilmore Girls Massyl Mallem (47:27): Are? No. Really? I feel like it rings a bell, but no, I've never seen an episode. But I will guess New Haven. Jared Correia (47:37): Good man. I like how you're just like, fuck it. I'm just going to guess. Now they lived in a made up town called Stars Hollow, Connecticut, which is not real. Massyl Mallem (47:46): Stars Hollow? Is it like a Harry Potter town? Jared Correia (47:49): Yeah, kind of. It's like a totally made up town. Okay. Never watched the Gilmore girls, so me neither. So don't feel bad Massyl Mallem (47:56): About that. Did catch a few episodes of Gossip Gold when I was a teenager. Really? Okay. Jared Correia (48:04): Do you want to hit me up with another contemporary reference that I won't understand before we finish up? Feel free. Massyl Mallem (48:10): No, no, you're all good. Wait, Gossip Girl is not even temporary. That's like 90s, right? Jared Correia (48:15): It's contemporary to me. Massyl Mallem (48:17): Okay. Again, sorry for. Jared Correia (48:19): Pop culture stopped for me in 1998. I had a good run. Massyl Mallem (48:24): Nice. Honestly, it only took a downhill from there probably. Jared Correia (48:29): Ms. Field, this is a lot of fun, man. Thank you for coming on. Massyl Mallem (48:33): Pleasure. Thank you for having me. Sorry for completely bombing your Jared Correia (48:37): Testimony. It's better when people totally get destroyed. Enjoy the Nutbex Massyl Mallem (48:41): Thing.s you win. Yeah. No, you win. Congratulations. I'm proud of you. You did great. Jared Correia (48:46): Oh, it's not a competition. In any event, come back again. We'll talk to you soon. Take care. Massyl Mallem (48:50): Pleasure. Thank you. Have a good Jared Correia (48:52): One. Thanks to our guest, Maciel Malim of Patent+AI. To learn more about Macsil and Patent+AI, visit patentplus.ai. Look at that curve ball. Patentplus.ai. I think we have our first AI domain name here. Exciting. Now, because I'll always be a '90s kid who would love nothing more than to taste Sobey Love Bus Brew. Just one more time, that's not as perverse as it sounds, trust me. But whose true passion is burning CDs for everyone who would listen. I'm not just doing the morning version of that, which is creating Spotify playlists for every podcast episode that I record where the songs are tangentially related to an episode topic, sometimes very tangentially. This week's playlist is chill as fuck. That's right. Homies, it's winter songs. It's here to warm your soul and is sponsored by Lucky Charms Hot Coco. That's a thing. They just haven't actually sponsored this show. (49:51): Curse you lucky. Join us next time when I learn about animal husbandry. Don't forget the rubber gloves you wankers.