James Pittman: Welcome to Immigration Uncovered, the Docketwise video podcast where we dive deep into the dynamic world of immigration law, shedding light on the latest developments, cutting edge practice management strategies, and the transformative impact of legal technology. I'm thrilled to be your host on this exciting journey as we empower immigration lawyers with able insights and explore the intricate intersection of law and society. I have as my guest today Amelie Sophie Vavrovsky. Amelie is the founder and CEO of a company called Formally. She is also a Codex fellow at Stanford Law School and an artificial intelligence researcher. Amelie, welcome. Thanks so much for joining us.

Amelie Vavrovsky: So great to be here, Jane.

James Pittman: So Amelie, to kick things off, could you share the inspiration behind founding Formally and your vision for its role in the legal tech landscape?

Amelie Vavrovsky: 100%. So, I think on an Immigration podcast, it's important to share that I am here on a visa. So I went through the immigration process myself and just noticed that even though I had fantastic counsel, it felt challenging at times as an immigrant, as a consumer, to figure out what am I eligible for, how do I actually effectively find a lawyer, how do I work with one? So I think going through that experience myself definitely brought some perspective to me about what it's like to be a client going through an immigration process. And then the very first thing we actually ever built was Formally was a tool to help people apply for asylum, like for Asylum Zone was to do remote asylum applications, especially during the pandemic. So that was an exciting time. But I think the insight, and we're going to take it to answer your question, James, is really around creating sort of a connective, creating the connective tissue between law firms and their clients, whether those clients are corporate clients or individual clients, and really giving them a modern, sleek experience and interface combined with the best counsel. So I think creating a notion that having the best user experience and the best client experience, in addition to the greatest attorneys, the greatest counsel, I think is the vision that we have for the future. So no trade offs, right? No need to just have AI products that replace lawyers and also no need to sort of hold on to interfaces that don't work for people who are non lawyers. So we want to combine sort of like cutting edge technology with the absolute best legal advice that people can get to create a better legal experience.

James Pittman: Well, you're really focusing on collaboration and collaboration is a key aspect of Formally. How does the platform facilitate collaboration between lawyers and their clients, especially in immigration cases, and especially if you're involving remote communication?

Amelie Vavrovsky: Yeah, so we've really reimagined sort of what the client portal experience looks like for immigration law firms specifically. So creating really seamless frictionless and AI enabled onboarding experiences for clients where it's really clear what sort of the expectations are what documents are required, and then we use AI to make the process a lot more seamless. But I think the classic client portal experience has sort of not been like the central focus of most technology that I've seen. Most technology focuses on giving the law firms a great experience, which is really, really important for efficiency and internal workflows at the law firm. But we're sort of adding on that consumer client experience layer on top of those technologies to make sure the technology doesn't just work for the law firms, but also works for both corporate and individual clients. And that's sort of how we're thinking about formally as a platform.

James Pittman: Understood.

James Pittman: Well, Amelie, your background is very impressive, I must say. And includes being a Codex fellow at Stanford Law School. First of all, what is a Codex Fellow and how has this experience influenced your approach to AI research in the legal domain?

Amelie Vavrovsky: That's a great question. So Codex is the center for Legal Technology at Stanford, and they've been doing legal tech research for a really long time. And that means there's sort of computational law, right? And then there's everything else that goes into legal technology. So they've been thinking about legal tech and AI in a really deep way for a long time. And I think the perspective that I bring that I'm really passionate about is what does it mean to create accessibility for non lawyers, right? Like consumers, clients in the legal space. So, again, focused on thinking about access to justice. What makes a legal workflow feel effortless, seamless, easy to somebody who's a non lawyer? What makes it easier to navigate a legal system and then also thinking deeply about the ethics of using AI, for instance. Right. So what is AI good at? What is it not so good at? I did a lot of research early on when there were sort of technologies in courts, like they were contemplating technologies to sort of decide what type of sentence someone should receive and what are the biases inherent to these systems and should we even let sort of an automated system decide or make decisions at that sort of scale. And where I err on the side of should we do that, should we not? Is probably not. Right. And I think the reason for that is that what we call AI, there's so many different subparts, whether that's machine learning or large language models, they don't really have self reflective capabilities and capacities. Right. They can't really reason about their own decisions in the way humans can. So if there is bias inherent in the data and the training data, there can be an effort to mitigate that, but it's likely going to reproduce it and not know why. And when you ask an AI system, why did you make that decision? Especially if it's a generative AI system, it'll just give you a response. generatively. Right. So the response will sound eloquent, but it might still not be true. And I think those are the interesting nuances. So to bring it all the way back, I love Codex, right, for sort of providing the research, the insights, like what's possible. But then I also like to play the role of translating that back into practice. Right. What does that mean for a law firm that can't spend their entire time thinking about the theory of large language models and machine learning? What does that mean practically for them? How can they leverage these tools in a way that feels safe and responsible for both their firm and their clients? And I think that's something I'm passionate about. Right. I love doing that translation work. I think you know that I'm also a huge language nerd, so I like to translate literal languages, but I also like to translate between disciplines. So thinking about computer science on one hand and then lawn policy on the.

James Pittman: Other hand and then was your background actually computer science from the beginning?

Amelie Vavrovsky: No, it's something I sort of took a few classes in. I'm going to be very honest with you, I can write some code. I'm not very good at it. I don't actually write code today. I have a really great engineering team that does that today. But even having the foundations and spending time sort of with a lot of computer science PhDs has sort of given me an appreciation for how they think and how computers think. Right. Which is, I think, a really valuable perspective to bring as you're thinking through policies for law firms, but also policies as like a nation. Right. How do we want to think about this new technology? How do we assess it? What are the right questions to ask?

James Pittman: Well, artificial intelligence obviously revolutionizing countless industries, but what specifically brought you to law?

Amelie Vavrovsky: I have always had a passion for sort of human rights and justice overall. I've had a very international upbringing, so I think what I initially wanted to dedicate my time to was international policy, international law, international human rights law. And then what inspired the sort of like tech angle to all of that is that this notion and maybe I'm getting a little bit philosophical here, right? But I think technology is so omnipresent in our lives and will continue to become more and more present that your rights online matter just as much as your rights offline. So if you don't have good frameworks to think about what that means right? Like what does it mean to have right to freedom of association or expression online just as much as offline? I think if we don't think about these rules, then I think we're going to not set ourselves up for success. So that's something I got really passionate about, is thinking about law policy from that technology perspective. But yeah, I've always been passionate about law. I actually thought it was going to be a diplomat, so it was international law, but I think then it just scaled back to human rights law. And I think immigration law is still a super international field of the law, right? So that sort of also made sense to me intuitively.

James Pittman: From what you're saying, access to justice is a critical aspect of what you're interested in. So how does formally contribute to bridging the gap in access to justice and ensure that more people have access to legal resources, particularly in immigration matters?

Amelie Vavrovsky: So we are still doing work with asylum seekers, for instance, right? So we're partnering with nonprofits to help provide them with great tools and great technology. But to get to what formally as a platform actually does, it's really about breaking down complex legal processes into digestible chunks that make sense to somebody who has no background, right? Like, no legal background and doesn't really understand the process and what it looks like. And I think there's a few sort of core tenets that I think about when I think about accessibility of legal systems. So one is actually just providing clarity on process, right? Like, what can I expect? And I think this is a very famous sort of thing in the legal world, is like this legal black box, right, where you don't really know. You work with an attorney and then you expect something to happen, but you don't really quite know what that is. So what does the process look like from beginning to end? I think providing that transparency is a helpful tool in terms of accessibility. Another helpful tool is actually providing translations, again, both into other languages, but also thinking about translating legal text to something that people can understand. Right? So if a USCIS forum asks you and that's the question on the asylum application, check this box if you also want to apply for withholding of removal under the Convention Against Torture, what does that mean? Right? Like, if you're applying pro se, is there a way to explain, to give enough context for this question without providing legal advice, but providing some context that might allow somebody to answer that question? And then I think there's maybe two more things that I think about. One is thinking about the emotional state that somebody is in when they are going through immigration. And I think that behavioral psychology is actually really important, right, because most people are really nervous and anxious when they go through immigration. And immigration lawyers know this better than anyone else, by the way. Right. Like, you need to really have empathy with the client and hold their hand through the process. So the way we design technology is to feel a little bit like that, right? And the initial pun for formally was like, it's like your ally for forms. We kind of moved away from the pun and we thought it was a little bit cringey in the end. But that was sort of the idea, right? Having somebody that can help you an ally that can help you understand what's going on. And then the last thing I think about, especially today, especially around AI, is delight, right? Can we make something that people are fundamentally finding scary or maybe clunky? Can we bring delight to that user experience? Can we make it feel seamless? Can we make it feel easy? Right? Those are sort of also things that we always look for as like those magic moments for client experience.

James Pittman: That's really very interesting, what you're talking about. So you're talking about leveraging the artificial intelligence and leveraging insights from behavioral science, behavioral psychology. What is it that brought you to sort of combine those elements and have them work together to improve the legal experience for what you've talked about it in the context of the clients. How is there any behavioral science sort of influence on your attorney facing side?

Amelie Vavrovsky: 100%. We did a lot of user interviews and that's where all of our insights come from. I think we very much try to take a beginner's mindset to all of our client conversations, right. You can't try to over rely on your own experience, right? Because your experience inevitably will not be everyone's experience. But as you start to sort of approach every conversation with curiosity, you start to gain insights and you start to see patterns, right? Like, oh, I was super nervous about this, and in my immigration process, this is something we hear from clients all the time. I just felt like I didn't know where to start, and then I procrastinated on it. And then you also start to hear that from the law firms, right? The law firms are like, my science just won't get back to me and I don't know why. And then you look into behavioral science and you notice, oh, anxious people procrastinate. That's what anxious people do. Right? Like anxiety. A response to anxiety is procrastination. So that then allows you to sort of tie those insights together. But yeah, our product designer has a behavioral psych background as well, so that also helps with tying those insights together. But yeah, that's sort of how we think about it all comes from our users both on the legal and on the client side.

James Pittman: So if the client's procrastinating, I mean, collaboration, the collaboration through the client portal is so central to your product. If the client is procrastinating and not responding, what are their automated features which then kick in and sort of prompt them.

Amelie Vavrovsky: Yeah, exactly. So we prompt them to come back. We also have something called a Task helper. This sounds like a really simple intervention, but again, we had this question, which is like, clients say, I just don't know where to start and I'm busy, right? I don't know if you've heard this, James, but I've definitely felt this with the client. I'm busy and I don't know where to start. And if I want to get started, I have to look through all my notes and then figure out what the right starting point is. So we introduced a task helper. So that just asks the client, like, how much time do you have today? And we'll get you something that you can do in five minutes. And if you have 15 minutes, we'll help you get things done in 15 minutes. Right, so even just that simple UI pattern, I think, can make a world of a difference for somebody who's just not sure where to begin.

James Pittman: Well, formally utilizes artificial intelligence, and you're heavily leveraging your research background in AI. So what are some of the specific AI powered features informally that can benefit immigration lawyers and their clients?

Amelie Vavrovsky: We of course, have features that are generative. Right. So helping write first drafts of letters. So thinking about sort of extraordinary ability petitions, which I am intimately familiar with, so helping people get through the first drafts. And we've implemented features on that side. And then we're constantly pushing the frontier of what's possible at terms of other functionalities. And we actually have an exciting announcement around a new functionality called Magic Docs. So stay tuned for that. James. We're going to launch it maybe actually right after this podcast drop, so maybe we can make something happen there. So we're excited to launch that publicly very soon as well.

James Pittman: All right, well, your platform also focuses on data security, and I know that's important to you. So can you explain how formally ensures the privacy and security of the sensitive client information that you're dealing with in immigration?

Amelie Vavrovsky: Yeah, compliant information is so important and so sensitive. So to nerd out here about data privacy a little bit, I did some work around privacy and privacy policy, but also kind of passionate about cybersecurity on the offensive cybersecurity side. So experimented with white hat security research and hacking and trying to really understand what's called like, a cyber kill chain. Right? That sounds really scary. All that means is, like, how do hackers or sort of intruders think about what's called escalating privilege? Right? So they try to find a vulnerability that might be like somebody clicking on a phishing email, and then how do they get access to more and more information? And what you notice when you do that kind of work and when you really look at cybersecurity research, you'll quickly notice that ransomware is on the rise. A lot of these sort of cyberattacks are on the rise, but most of the bad stuff on the Internet happens for really simple user behavior reasons, right. So somebody clicked a phishing email or somebody reused their passwords over and over again. So we look at those trends and we look at sort of like, what are the most important things to actually get? Right? So we do things like implement multifactor authentication. That's a really great way to sort of do that at the beginning. But then the other interesting thing we do is we actually encrypt data at rest individually in each record. And data for companies, for instance, company clients and employees, is kept separately. So all data is individually encrypted and at rest and is also portable for the person whose data actually is. So that kind of flips the script a little bit, right? Like as a client, you don't just email data if you use formally, you have your data on formally and you can share it with your law firm and we can share it for record keeping, but the primary owner of the data is actually the organization or person whose information it is. So I hope that makes sense. But that means that it's harder to escalate privilege, right, if you don't have super admin access and if you then have some gates around it. So we try to take those steps and we try to also make sure that users are actually aware where their data goes, who they shared it with, right, getting consent, being able to take it with them. Those are all privacy forward practices that we're all passionate about.

James Pittman: Well, you've hinted at the Magic Docs functionality, which you're going to be announcing and previewing, but just more generally, I mean, clearly you have placed a lot of emphasis on intelligent onboarding for clients and that is really one of the strongest aspects of formally. So can we drill down a little bit? How does the intelligent onboarding streamline the process of collecting information and preparing immigration cases?

Amelie Vavrovsky: You know what, James? Well, we'll talk about magic, Docs. If the people are listening to this podcast, I feel like they deserve a sneak peek of what's coming. But we will be dropping an official announcement. So the way Magic Docs works and the way onboarding works informally is that, again, the way to think about our product is that we have smart portable data platform that you can overlay, great, easy to follow workflows on top of, and those workflows are AI enabled. And right now, because we're so focused on immigration law, they're all immigration processes. So we actually did a lot of work with lots of different law firms to create really good baseline templates for say, an H or an O, right? Those templates are still customizable, but we already pre built them out. So what that means is that we built a client experience around an A, for instance, right? So what you can do as a law firm is you can customize the onboarding process. You can either select a template or not. You can also onboard somebody without a template. You can select what documents you want and what information you want to collect. And then we send an onboarding email, sort of one simple link to the client. They then log in, they sort of see the questions. And then here's where Magic Docs comes in and that's really exciting. They'll see sort of a list of documents that the law firm requests from them. And then here's another insight about users. Right. Again, I know this, all immigrants I know have this folder on their desktop. It's called immigration, docs. It's not organized and it's not pretty and it has like screenshots of old I things that aren't really labeled properly. So what you can do with Magic Docs is you can actually select all of those documents, right? Completely disorganized, complete disarray. And you can just drop them in Magic Docs. And what we'll do is we'll actually go through, we'll identify the type of document, right? Like we'll scan it and we'll see what type of document it is. We'll correctly relabel it. So if it's a passport, we'll label it as a passport and then the name that the client gave. And then we'll also pull out all of the information from the documents, right? So we'll see is your first name the same on your passport as it is on your birth certificate? And we'll show you where we got it from. So you are always in control, you always know where the sort of machine is taking things from, but it's a really magical experience and it really saves a lot of time and users are really excited about it. So I'm excited to show you sort of the demo of what that looks like. But yeah, it's a really exciting functionality and that makes the onboarding process really easy. And then once all your data is on formally, you can put that into forms, you can wrap that into full petitions, you can use that as input for generating letters, right? So you can do so much with it in any process that you sort of want to add on top of it. So it's a very different experience, but that's sort of what's coming on the onboarding.

James Pittman: Amazing. I'm sure everyone is looking forward to really seeing the full scope of that functionality when you make your formal announcement. I certainly am.

James Pittman: Amali can you share a success story? I mean, you don't have to give identifying details, but a success story or a case study where you've seen formally make a profound difference in a law firm's efficiency?

Amelie Vavrovsky: Well, actually, first of all, I think I want to actually question the notion of efficiency being the most important thing. The way I see is that law firms actually know their processes and internal processes best, right? So they actually are already efficient as is. I think where we actually have seen the and that was our initial notion, James. We were like, we're going to make it more efficient. And then I think we learned from our client that we were wrong, that efficiency was not actually the most important reason for a lot of firms to adopt technology. It was actually to gain and retain new clients, right? So what we're hearing from law firms that we're working with is like, having great technology and great client experience helps us win RFPs. It helps us get and retain new clients. Right? So I think that's actually where our most important success stories lie in terms of efficiency. We also have seen increases there mostly in terms of clients responsiveness, client procrastination, client communication, initial onboarding processes. But I think the biggest success stories are around having this technology in my portfolio helped me win a large contract with a corporate client that really wanted a modern portal that leverages AI or from company clients that were like, yeah, wow, onboarding was so easy. I didn't know that law could feel so easy and then just being super satisfied with their immigration legal experience. So I think those are the success stories and we're excited to uncover more, but I think we had to learn that was a humbling moment for me. Right. Like efficiency is not actually the only thing that matters because the best law firms that I know deeply care about giving the best counsel, even if that means that some processes are not the most efficient, such as like taking the extra time to hop on a phone with a client when they feel stuck. Right. That maybe is not the most efficient thing to do, but it's the right thing to do and it's the thing that will provide long lasting, high value client attorney relationships. So that was a big learning for us, is to really trust the law firms in their processes and really hone on how can we help you do what you already do best and scale that to your clients.

James Pittman: Well, let's probe the communication aspect of it that you've just sort of mentioned. So client communication is so very important and pivotal in practice. How does formally ensure effective and transparent communication between lawyers and clients during immigration cases?

Amelie Vavrovsky: I think it lays it out in one place, right? It just makes it really clear like what is asked of and what is required. So the communication moves sort of from a more logistical focus to more of a substantive focus. Right. So what we've noticed is a lot of initial onboarding conversations are getting the basics down, right? Like I have a question about this particular question in the questionnaire, or I'm not sure about this criterion. So I think formally proactively answers a lot of the questions around process, around what's required, which leaves more time to focus on the substance of the case or the strategy around framing a particular case. So I think it's just shifted that focus from more logistical. And I think again, what we hear from law firms is like, we get so many emails being like, where's my case at? When are we going to file? Right? And I think those all come from a place of like, I don't actually understand the full process and those are the types of communications that we can dramatically cut down on. And then on the legal side, we have some interesting functionality to make sure that law firms sort of know where their clients is at. And we have really granular notification settings that give law firms a little bit more control back over their inbox, right? So instead of getting inundated with individual emails, law firms can sort of choose to set notifications to be batched, for instance, right? So they can say, I want to get a weekly update, a weekly summary of everything that's happened in terms of document uploads, or I want to get a weekly summary of all the client notifications. But you can also choose to get immediate notifications by email or text. So that's how we think about it. I hope that answers your question.

James Pittman: Yes, absolutely.

James Pittman: Now, we've talked about various aspects of the AI and some of the AI powered features informally but sort of speaking holistically getting from where we are now to sort of the next stage. What role do you see AI playing in shaping the future of, first of all, immigration law practice and beyond that legal practice as a whole and how is formally positioned to lead or stay in the forefront in this regard?

Amelie Vavrovsky: Here's what I think is most important. Knowing about generative AI, for instance, right? I think the thing that I really want to focus on is getting everyone on the same page in terms of what GPT does and does not do, right? So I like to get sort of to the foundations and the basic, which is that fundamentally GPT is a probabilistic model that generates words, generatively, right? Like word for word. And that's kind of a crazy thing. It's so crazy that it works, right? But it also makes you understand why hallucinations can happen because it's all probabilistic sort of generation of words, right? So what I'm trying to say here is once you actually dig into the technology and you understand what it can do and what it can't do, you'll also be less surprised, right? So what we hear from law firms, Francis, is like, GPT is hallucinate and chat GPT hallucinates. And I think that that's not only unsurprising, but also very probable and likely. So I don't actually see GPT or large language models or like generative AI as a tool to replace legal drafting and writing. What it can help with is like, first draft, especially if those drafts are generated by non lawyers, right? So if I need a letter of support for my O, i, as the client, need to be able to understand, I will be able to verify whether this letter is actually true or not true because it's not fundamentally a legal argument, right? It's just like, this is what I accomplished. So as a layperson, I can easily identify like, okay, this information is correct, so it can help tremendously in speeding up the drafting process. But if I need to make a legal argument, you will always need a human lawyer to sort of verify and check that information. And that, I think, is really key to. Understand and also helping law firms understand and setting expectations of this will provide you with a first draft that sounds eloquent but really requires fact checking and nuance and verification, I think is really, really important. So that's the first point, right? It's like, can we bridge the information asymmetry gap, right? Can we make sure that we're all on the same page around what this technology is, how it works, so that we can ask better questions and make better decisions around it? And then in terms of opportunity, the way we think about it is like, everywhere. It can create great user experiences and customer delight. That's where we're most excited about it, right? Can we create a seamless drafting process that gets you over writer's block? Can we help you extract data from documents and cross check it and double verify it so that you don't have to manually label your documents and manually do data entry? Right? Like, nobody really likes to do that kind of task. And that sort of scalable repeatable task is actually something that AI and technology overall is really, really good at. So we're going to keep pushing the frontier of what's possible in terms of repeatable tasks. But we're also I feel extremely clear in my mind that lawyers have a really critical and important role to play in terms of being of counsel and helping people navigate nuanced legal situations in a way that AI systems simply cannot because they don't have reasoning in the way that humans do, especially not well trained attorneys. So that's my prediction and those are my two cent on it. But I think if law firms can learn how to collaborate with AI tools, then I think that's probably the ultimate winning combination over the next few years.

James Pittman: Yeah, no, those are definitely very valuable insights. Thanks for that. That's very enlightening.

Amelie Vavrovsky: Let's go back a little bit to.

James Pittman: Your interest in human rights and the intersection between artificial intelligence and how it impacts law and human rights. So what sort of ethical considerations do you think that legal tech developers and entrepreneurs should keep in mind when you're designing AI solutions for legal practice?

Amelie Vavrovsky: My biggest concern around AI tools for access to justice that I think needs to be thoughtfully designed and considered is that if you think about the people who most need legal services, right, who don't currently have access to them. And you also think about people who would have the ability to critically evaluate information and assess whether it is truthful and accurate and have the understanding of how it was generated. I think you'll notice that the people who are least able to vet the information are also the people who most likely need legal services and great legal counsel. And that's really important to acknowledge. There's a disproportionate negative impact of giving false generated, false legal counsel to people who are most in need. And I think there the repercussions cannot be like the severity of those repercussions cannot be overstated. So I would be very careful trying to give any form of legal advice to individuals who don't have counsel on the other side. And I think that that's very much the way this information works on the Internet. Right? We know that you shouldn't believe everything on the Internet, but if it sounds really credible and true, maybe you'll believe it, right? Because it sounds like an authoritative source or it looks like a good source and it's hard to know exactly how to critically evaluate sources. That's something you have to actively learn and acquire as a skill and have the fortune to have a school or education system that teaches you that. So what's going to happen with generative AI specifically, is that the spread and quantity of eloquent sounding disinformation is going to dramatically increase. And I think that's also going to hit the legal field by storm, right? Like, you're going to see a lot of advice articles that sound really smart, right? DPT sounds really smart, but might be wrong. So I think those are important ethical considerations. So as you design any sort of AI for end consumers, I think you should be really careful to draw clear lines around what it can and cannot do and making sure that it's not actually giving legal advice. Again, drafting something like a personal statement or a personal narrative, I think is a great use for AI, because a layperson with no legal background and no ability to vet a legal argument can clearly assess that information to be true and accurate. But if you need legal judgment, AI is not the right place. That's my take on it, and that's how I think about it. So those are some of the ethics. And then I think there's also you need to make sure to, for instance, check that all data that goes through the system is not used for training models after the fact right. So that you don't have rude awakenings to models being trained on your data. So, checking, is this actually used to train more data or is it not used to train more data? I'm actually working on this, James. I'm working on a guide for law firms around what is GPT and how to ask the right questions around it. So that's, I think, something I'm excited about doing more of, but like, those are some of the ethics considerations and how to think about them. Obviously, this is the tip of the iceberg, but the general takeaway here is be really careful to not promise that GBT can give legal advice, especially to people who are in need because they're most likely disproportionately negatively affected by it. And then also make sure to ask your providers the right questions around what happens to your data and how it's being used to train more models.

James Pittman: Yeah, it's very thought provoking. I mean, to think that immigration lawyers have spent a lot of time sort of trying to guard against clients and vulnerable populations, sort of being given bad advice by notarios or unqualified people, sort of in the immigration space. Who are not attorneys or not trained to think that you will also now going to need to guard against the AI equivalent of those actors. Yeah, that is quite thought provoking. Definitely.

James Pittman: Well, Amelie, I've seen that you've mentioned that data equity is one of your focus areas. First of all, what do we mean by data equity? And how does the creation of formally help to address data equity concerns in the context of legal technology?

Amelie Vavrovsky: The way I think about it is that as a modern consumer, our digital identity matters, right? It matters where your information goes, it matters where it's stored, it matters who has access to it. Right. So understanding where your data goes, who it's shared with, how it's handled I think all of those things would fall under the umbrella term of data equity in my mind, just making sure that people have clear understandings of their personal information, where it goes, and that they have access to it in a way that makes sense to them. Right, so building that accessibility layer as well. So empowering people with their own information, I think is really powerful and I think you can do really cool things with it too, right? Like, once you empower people with their information, you could think about making smart recommendations, for instance, right? Like, based on your past filing for an asylum application, we think you're now eligible for an employment authorization renewal and a fee waiver. Like schedule an appointment with your council, with your nonprofit to check in on it. Right. So you can make proactive suggestions and help people sort of in the longer term. So, again, it's all about data. My team makes fun of me. They're like, this is Amelie's data dream. But I think having smart networked data that serves clients and sort of their journey, their legal journey, is really powerful and something that's going to be really big over the next few years.

James Pittman: Beyond immigration law, are there other areas of legal practice that can benefit from the technologies and features offered by formally? And have you begun thinking about expanding your services to other than immigration law firms?

Amelie Vavrovsky: 100%, I think that the technology and we get a lot of requests for it. I think we get a lot of requests from especially law firms that have multiple practice areas. I think the technology is very scalable. But right now our focus is on immigration law. And I want to be very clear that we want to be judicious about how we think about expansion. Right. Because we don't want to be a generalizable tool that doesn't really work for anybody. Right. I think that there's a risk that you run when you try to be everything to everyone. Like, you run a risk of being like nothing to everyone. So we're really focused on immigration, and we really want to make sure that when we work with immigration law firms, we do the best possible job. Like, we want to win the hearts and minds of immigration law firms first before we expand. But there are some interesting and exciting expansion plans on the horizon, for sure.

James Pittman: Well, Formally and the client collaboration and client portal that you're talking about and the Magic Docs functionality are very advanced tools.

James Pittman: When you encounter lawyers and law firms, I mean, you must encounter some that are sort of behind the curve with technology. How can law firms incorporate a tool like Formally and AI generally and other tech solutions into their practices without feeling overwhelmed by the technology? What's your advice regarding implementation?

Amelie Vavrovsky: We've heard that implementation is easy and seamless so far. That being said, it is completely my responsibility and our job as a company to make sure that that's easy, even for a firm that hasn't really played with technology all that much. Right. So I think it's on the provider company to make sure that we provide you with a great experience. So when we work with firms, we come in with a three phase implementation plan, making sure that we have a way to get you comfortable with the tool. And then at the end of the day, Formally really is very straightforward on the legal side and really designed to be comprehensive on the client side. So I think a lot of the functionality is actually client facing, and there we continue to get really positive feedback from both companies and individuals that go through the process. But when we get feedback, we take it really seriously. And I am not joking. When we get feedback, it's a gift, right? We will send our clients flowers. We're really excited when you give us feedback, even if it's negative feedback. So I think for firms that haven't played with technology, there's extra value for us because we get to learn about their perspective from sort of a new lens. Right? Like, what is it like to implement a system for the first time if you haven't done that before? So we're excited to take down that challenge and we're excited to get even more feedback from law firms who have less experience implementing technology so far.

James Pittman: Speaking in a general sense and even beyond immigration law, given your insights, what are the larger trends that you foresee shaping the future of legal practice? I mean, we've talked about the generative AI, but other sort of large technological trends that you think are going to have the most impact.

Amelie Vavrovsky: Yeah, I tend to look at other industries to sort of understand what the differences are between legal and other industries. Right. So I look at Fintech, for instance. I think that one of the trends is that client experience again, I keep saying this. I feel like I sound like a broken record, but client experience is really paramount because it means that you get to retain and gain new client contracts over and over again. So I think there will be a trend around creating better and better experiences for people going through legal processes. I think that so far, law has been relatively untouched by sort of dashboards and platforms that feel really seamless on the consumer side. So I think that's one trend. I think that drafting, obviously with AI is going to be a big trend. I think that law firms are going to start to even more critically look at privacy and data security. So that's another trend I would predict is that the modern consumer cares deeply about their data, especially the corporate clients specifically care about data. So I think that's another trend that I see in our future is people asking more and more questions around security and giving more security questionnaires even to their law firm partners. Yeah, I think those are some of the major trends, right? It's around data privacy and security. But then again, I have a really particular point of view. But I think, yeah, law firms are going to enhance processes, do a lot of process automation, and I think some of the work that law firms do might shift a little bit, right? So I think paralegals and legal assistants will likely be leveraged in new and creative ways to build client relationships, to do sort of outbound marketing, to help with other tasks. So I think that the law firm structure might change a little bit, and I think that lawyers will do even less work that feels like manual, right? Like there's probably going to be a lot less manual verification and a lot more collaboration with technology and hopefully a lot more FaceTime with clients.

James Pittman: Well, Amli, your background includes an interest in digital rights, and we've talked about data equity. Now let's talk about the concept of digital rights. I mean, what do we mean by this term and is that different from data equity? And how does formally ensure that clients data rights and privacy are protected in alignment with data rights principles?

Amelie Vavrovsky: What I mean with digital rights, or what the field really looks at is I think we touched on it a little bit at the beginning, right? But what does it mean to exercise your rights in the digital space and how does that play out online? So thinking about I did a lot of research that is actually not legal or immigration specific, right? But thinking, for instance, about content moderation policies on social media platforms and thinking about who's allowed to post things, what types of posts get taken down and why they get taken down and are there political reasons and who makes the moderation policy decisions, right? Like, how does freedom of expression play out? But then, even if you think about things like imagine this scenario, James, like if you had a crowd of protesters on Market Street in san Francisco and they're marching for a particular cause. And then you have drones with facial recognition technology flying over the protesters. So suddenly, theoretically, you could know who was at the protest even though they were not posting about it online. Right. So the online and the offline space are suddenly converging in a new way that probably makes us, I think we should think now about what that means. Right. Especially in the hands of an oppressive regime, for instance. So those are all parts of digital rights and those are some of the work I did in Middle Eastern countries. Right. Thinking about what content moderation policy should look like, what does it mean to have incitement to violence online, when should police forces intervene? So I've written a lot of papers and done a lot of research on those exact topics to make sure that when those things happen. And I think it's just a matter of time until we encounter more and more of these complex issues. Many of them we're already finding. How do we handle them, what's the policy blueprint, how do we think about them to bring it back around? Like your question around house formally aiding with that data, I think it's around just giving people transparency into their data record and their legal records and their transparent relationships. Right. So instead of just having data on law firm servers and then law firms having to go through a whole process to sort of provide that data back to the client, what if we made data portable and automate reporting for law firms? Right. So that for reporting requirements, they have all the data they need. But how do we make sure that that's a consensual experience that feels seamless and easy for both the law firm who doesn't have to respond to these requests for information and data now, and also the consumer who then knows exactly where it went, who had access to it, and how it was handled.

James Pittman: Understood. Very insightful.

James Pittman: So Amelie, you were a Forbes 30 under 30 honoree in legal tech. And as an entrepreneur, what advice do you have for aspiring innovators who are trying to make their mark in the legal tech industry? And also could you share some of your most impactful lessons that you learned on your journey as an innovator in the tech space?

Amelie Vavrovsky: I think the most important thing is to really focus on a problem rather than a particular solution when you approach any innovative process right. Like, what are you trying to solve for, what are your hypotheses? And then assume that you're wrong. Assume that you're probably wrong about your hypotheses and get really systematic about sort of proving yourself wrong over and over again until you find something that feels right and looks right. And if you have a passion and a fire for the problem you're trying to solve, you won't get as discouraged by the solution changing. Right. That's kind of a normal side effect. If you are focused on the problem and not attached to your solution, then things can change. And I think that's exactly what you want at the initial phases and stages of innovation. So find something you care about and then find a way to test hypotheses over and over again until you find a solution that actually solves the problem that you're looking at. Would be, I think, my biggest piece of advice in terms of lessons, I learned so many, but I think, yeah, it's okay to prove yourself wrong, right? Have humility and an open mind every time you go into a customer conversation, be ready to learn something new and have that growth mindset, right? Like every time you learn something that is a gift, that is helpful, that helps you get closer to the right solution. And I think once you have that mindset, even challenges like we had that bank meltdown, you get so much better at navigating those things if you can see the value and the learning of what you got from it, right? So from the SVB meltdown, we learned about FDIC insurance and how banks work, all of those mechanisms. So I think if you have that mindset, that helps you through all of the ups and downs that are kind of inevitable when you are on a startup journey. And then the last piece that I want to make sure I really emphasize that. Find yourself a great team, right? There's no way you can do it alone. I think the way I think about hiring, I love hiring because it's like assembling the Avengers, right? I feel like I'm doing a call to get the top people on our team. So make sure you surround yourself with top notch people who are all better at something than you are. That's actually really key. I think that's a really important thing I learned. You don't want to hire people who are worse at things than you are, right. You want to make sure that you have the humility and the excitement to find people who have different skill sets, who can contribute a new set of skills and that you can learn from as their leader. And I think that creates really strong working relationships and that's something I love most about being a founder.

James Pittman: That's golden advice. It's very important to have people who have complimentary strengths. So there you have it for all of our startup people. Golden advice from a rising star here.

James Pittman: So Amelie, it's been great and we're getting to the end of the hour, but is there anything else you want to tell our audience? Anything to legal professionals who are very interested in the topic but are sort of wondering where to start with leveraging AI?

Amelie Vavrovsky: Yeah, I think play with it, right? Get sort of into it, try it out. We are expanding our beta, so we are excited to have conversations with any law firms that are interested in giving it a shot. And don't be afraid. I think AI feels like a complicated new technology, but we've done this over and over again, right? We've had to look at the Internet. What does the Internet do? What does it not do? Just have that beginner's mindset as you approach the technology and just play with it and try to understand it better. And I promise that it's not as complicated or scary as it might seem at the beginning. So dive right in, give it a shot. Hit me up if I can help you. I love talking to people about this stuff. It's a passion of mine. And, yeah, don't feel discouraged by sort of buzzy words that often don't mean all that much.

James Pittman: There you have it, ladies and gentlemen, Amelie Sophie Vavrowski, founder and CEO of formally artificial intelligence expert and Codex fellow at Stanford Law School. It's been a pleasure, Amelie. I've really enjoyed the conversation. Thanks so much again.

Amelie Vavrovsky: Thank you so much, James. It was so fun to talk to.

James Pittman: You and join us next time for our next episode of Immigration Uncovered, the Docket Wise video podcast.

‍