James Pittman: Welcome to Immigration Uncovered, the docketwise video podcast. I'm your host, James Pittman, and this is episode 33. We're joined today by Khalil Zalawi. He is the founder and CEO of CasePoint. Khalil, welcome to the program.

Khalil Zlaoui: Hi, James. Thank you so much for having me. I'm very excited to be

James Pittman: on your broadcast. You're very welcome. Well, Khalil, to kick things off, could you tell us about Caseblink share share about the inspiration behind the company and your vision for its role in the legal tech landscape?

Khalil Zlaoui: Yes. Absolutely. So maybe just to kick things off, I'll tell you what CaseLink is about and who we are. So CaseLink is AI software that streamlines writing intensive, workflows and immigration. And we're starting with, O one a's, PB two national interest waivers, and PB one a's. We are a team of AI scientists, lawyers, psychologists, really with the right expertise to be building this product, and we have, a couple dozen customers using our app already, and they're have been they have been very satisfied with what caselink does. In terms of the inspiration, so I'm an immigrant myself, and I came to the US a couple of years ago as a graduate student. And, you know, I have been through the immigration process myself, applied to any sort of visa that you can think of from an f one to, the STEM OPT, h one, h one b and then transferred my h one b and then an o one a and then a national interest waiver. So I know how complex this process is. And just to tie that with really what inspired me to build this product for immigration lawyers is that my training is in data science and machine learning. And at the time I was working at Google building AI software for physicians and specifically on top of clinical notes data, so a lot of text data. And we were essentially digesting that for physicians to be able to do their work faster and not having to, you know, spend a ton of time reading clinical notes and sort of just like getting insights direct directly on the platform and summarizing all of that information for them. And I thought, we can also do this for immigration lawyers, and that's how CaseBank was born.

James Pittman: And where were you from originally? I'm curious.

Khalil Zlaoui: I grew up in Tunisia. And then after graduating high school, I moved to France and I spent almost 10 years there before moving to the US.

James Pittman: Alright. So now your background is very impressive, and it includes being a data scientist at Google. What was that like, and how has that experience influenced your approach to AI research in the legal domain? It was a greatest experience, especially working with, you know, the most talented people in the

Khalil Zlaoui: industry, building, sophisticated products. And, you know, I very much was building a product in AI, and so this definitely has, inspired me in terms of how we are building Case blank. And specifically, the philosophy at Google was to build and iterate and, you know, build, break things, iterate, and repeat. And that's how you get to, very, efficient AI models that truly understand the tasks that they have to perform and get to good quality products. So you have to have humans involved in the loop, engineers, data scientists, product managers, and that's the falls philosophy that we're trying to replicate, at CasePoint.

James Pittman: And but, I mean, Google is, of course, a giant company, and CasePoint is a start up. So in terms of trying to utilize that same methodology, I mean, are there any any issues or challenges with doing so given the the difference between the start up and the established company environment?

Khalil Zlaoui: Definitely. So although we were kind of in the start up environment there because it it was, for this very specific new product that we were building, It was definitely a much larger team. I guess what I was trying to say is replicating that philosophy in terms of, you know, hiring people at CaseLink and who the the these first key hires should be. So, we have a person who's very strong in engineering and infrastructure, someone who really knows about immigration. So we have an immigration lawyer on the team as well as part of the founders, and then we have ML scientists and, machine learning engineers. And so really just having those key players to be able to build even, you know, very great first versions of this product.

James Pittman: Yes. Okay. Well, let's dig into CaseBlank a bit. So CaseBlink, as I understand it, aims to make immigration legal workflows more efficient using AI. So can you give us an overview of how your platform accomplishes this and its potential impact on immigration lawyers' practices, let's say, from an operational and a business perspective?

Khalil Zlaoui: Yes. So I would start with the latter piece, so from an operational and business perspective. So we're starting with, again, these writing intensive work flows, the, 0 one a's, EB one a's, national interest waivers. They are very complex and they take a lot of time. And so because immigration is flat fee, by saving lawyers 80% of their time spent on these workflows, it really enables them to either take on more cases or make more margins. And so, it's it's a no brainer in terms of needs for automation and for these, type of work. What we do at Caseblink specifically for these writing intensive workflows is based on 3 pillars. The first pillar I call document understanding, the second pillar is case research, and the third pillar is drafting or tailored drafting. So document understanding is really understanding the profile of the candidate that who is applying to, the specific visa, what information to extract from based on those documents. So we train AI to do that automatically. It knows, you know, whether a piece of information coming from a document, it needs to go and, into, you know, a form or it needs to go into some some place in the legal brief, or maybe it needs additional research. In the case of additional research, we are building that pillar as well. So the AI that we build runs that with that information that is extracted from the document understanding piece and then performs additional research, whether it's on the web and our internal databases that we're building to support, you know, though those workflows specifically. And then finally the drafting piece. So the drafting piece has to be done very meticulously and carefully because we know the challenges with large language models and that they can hallucinate. So we really try to give the right context to be able to put together briefs that are very precise and that are based on the 2 first steps that I mentioned and also tailored to the visa that is being, considered. So if if it's an O and a, then it would speak to in in in the terms, an O and a. If it's an e b two national interest waiver, then it would be tailored to the national interest.

James Pittman: Now immigration law is a a complex statutory scheme. Right? It's it's known for its complexity. People always say it's right up there with the tax code. So how does Case blank assist immigration lawyers with legal research?

Khalil Zlaoui: As you say, it's very complex, and the way we're approaching case research right now is very much tailored to the workflows that we're automating. So we see ourselves as an end to end platform for those specific visas. And so the case research that we do is very specific again to those writing intensive workflows. So in the context of a national interest waiver, for instance, we kick off a few searches that are perhaps related to the candidate, their profile, additional searches that would, make their case stronger. We would look into things like executive orders, the CET list, and all of that we automate through AI.

James Pittman: Collaboration is an important aspect, and law firm staff have to work together and with the clients. So does Case blank have features that allow collaboration among the lawyers to staff and the clients, you know, especially when you're talking about dealing with people remotely, which, of course, working remotely with clients is is a typical practice nowadays, but are there case link specific features that facilitate that collaboration?

Khalil Zlaoui: Yeah. So I want to start by saying that, lawyers are already using a lot of great tools for communication, and there is already, you know, great CRM out there, and, you know, standard tools like Zoom, which, we know that the usage of those tools has very much increased since the pandemic. At CaseBlank, we're really focused on, you know, solving the pain point that the lawyers are experiencing right now with these workflows, these writing intensive workflows. So the focus is really to, save them time on those workflows specifically. Now we are inherently going to improve communication between the client and, the lawyer because we are essentially building platforms where, a lot of those steps are pretty much outlined in the software.

Khalil Zlaoui: And so it takes away that back that need for back and forth, interaction with the lawyer because everything is so well streamlined on the platform that a lot of what is lost in translation gets solved, by building pipelines and, workflows that are very clear both to the lawyer and the client.

James Pittman: So the cornerstone of the of your product is artificial intelligence. So let's talk about some you you mentioned a few, but are there any other that we didn't hit specific AI powered features, where that really harness the AI power for immigration lawyers and their clients?

Khalil Zlaoui: Well, without revealing too much, I will tell you what our overarching goal is. The overarching goal is to generate a finalized packet pretty much at the click of a button. So I know it sounds crazy right now, but I believe that in a couple of months, this technology will be able to help with tool automation. And so we're really talking about, you know, ingesting documents, understanding, the missing context, augmenting those documents with the missing context and the research that needs to be done, and finally, producing a finalized packet, which would include, you know, expert letters, exhibit lists, cover letters, and legal briefs.

James Pittman: So is there actual onboard when you're using CaseBlank, is there actual onboarding for clients, I mean, where they they're entering data and submitting documents through through CaseLink? And if so, let's, with I mean, you know, I'll have to repeat, but to what extent does AI assist the onboarding process? And could you just encapsulate, you know, how does this feature streamline the process of gathering information and preparing the cases?

Khalil Zlaoui: Yeah. So that very much ties up to the document understanding piece where a lot of the manual intake is taken away by AI being able to understand the type of document that is uploading, what being uploaded, what to do with it next, and how that fits into finalized elements in in the packet, whether it's a brief or it's a form. In terms of onboarding, we give the law firm the option to have the client log into the platform and do some of that uploading work on their side. But we've also heard from a lot of law firms that they are really cognizant about their client time and they don't want their client to be doing a lot of work. And so we give that automation, we we provide the law firm with the automation capability so they don't necessarily have to, involve the client too much in in those intakes. So, whatever the client, you know, shared with them, let's say, on the Google Drive, they would be able to process on their side of the platform without too many pain points or, spending too much time doing that.

James Pittman: Alright. So let's talk about, a little bit more about your background because, again, it's very impressive. And you were also a a data scientist at CVS. What are some of key insights you gained about machine learning when you were there?

Khalil Zlaoui: So I was working closely with the chief medical officer at CVS Health and working on projects for leadership there. And what I realized is that, ultimately what matters is the the message or the, you know, what whatever will help them make decisions as opposed to the details of the models or why you decided to pick this model over this one. And more importantly, there's always an uncertainty component associate associated with, you know, machine learning, statistics, data science. And so you really also need to communicate that, that any sort of finding that you have there, it has to be caveated with the uncertainty. And so tying this to back to case blank, I think this really has me thinking about, you know, these are essentially statistical models. And so because they are statistical models, they can fail. They can predict things that don't necessarily make sense. And so, essentially, thinking about all the guardrails that we can build, to have AI that does not hallucinate, and if it does hallucinate being able to detect that and to flag that. I will add to that, that, you know, my most recent experience was in healthcare technology. And so we're very much aware of the importance of data privacy and data security and also are transferring some of that knowledge in terms of how we're building case Blink.

James Pittman: Okay. So you did sort of say that it was your and it was really your own experience with the immigration system that got you to make the shift over to this sort of space?

Khalil Zlaoui: Correct. And just the advent of generative AI and knowing that there's an opportunity for this space that wasn't there, you know, 2 years ago. I know that with the advent of new technology, these these are very complex workflows. Like, you know, we talked about this. It's very difficult to automate. But because the technology is evolving so much, I believe that today and in the future, these will we will be able to automate these end to end.

James Pittman: Well, I saw from, your postings, on LinkedIn that you just came off of participating in the entrepreneurs roundtable accelerator, there in New York City. So what was what was that experience like? I mean, what goes on in the program, and what did you gain from it? Do you recommend it?

Khalil Zlaoui: Absolutely. So maybe for your audience, I'll, explain a little bit more what ERA is. So ERA is Entrepreneurs Roundtable Accelerator. It's New York City's leading technology startup accelerator. It helped us tremendously in various aspects of how we're building a company, how do we go to market, who we should be targeting in terms of, you know, customers, understanding our market a little bit better, and as well as, you know, a a lot of advice in in terms of sales and marketing strategy. So that was tremendously helpful because, you know, we are exposed to founders who have been in our shoes before, and we talk to a lot of these people who understand, you know, the process that we're going through and what it takes to build a company from scratch. And, just to give you an example, one of our lead mentors in this program is someone who started a legal tech company. And so he knows exactly the, you know, the stages and is able to provide us the right advice at the right time. So definitely a great program and I think, definitely helps helped us accelerate as well. And I don't think we would have been here without ERA, at this point.

James Pittman: Could you share a a success story or a case study, if you have one, where CaseBlank made a significant difference in, immigration law firms' efficiency?

Khalil Zlaoui: I do. About 2 months ago, we we were actually wrapping up the national interest waiver workflow, and we we had started working on the 01a. We were almost there. And then this client comes to us and says, hey, guys. I know that you have NIWs available, but, one of my clients came to me and they want to file their 01 a before the USCISP increase. So basically, that left us with 1 week to produce that case. We were able to turn that over in 3 days, and, our the lawyer kind was able to file the file the petition, on time, and it got approved in 2 days. And I can tell you that Kind was extremely happy. And she said, I'm so glad I found you guys, and you you really took away that stress for me having to file in such a short, notice.

James Pittman: Hopefully, you put a nice testimonial on your website.

Khalil Zlaoui: We're working on that.

James Pittman: So let's talk about, a a serious and important topic, which is data security, and that's, obviously very important nowadays to all kinds of businesses. But as immigration lawyers, especially because of the quantity of sensitive data that they need to collect in order to complete immigration forms. So can you talk about Caseblink's data security and retention policies and how you ensure the privacy and the security of sensitive client information?

Khalil Zlaoui: Absolutely. And before answering that that question, I will tell you, again, most of, my experience was in healthcare technology and PHI data is even more sensitive than PII data in healthcare. And so I'm used to, you know, data protection, data sensitivity for, in, in the space of healthcare. And I can tell you, we're very cognizant about data security at Casemink as well. Not only me, my, second co founder, John, our CTO, spent more than 15 years building health, startups. So we really know how to, you know, take these into consideration and, build safe and secure systems. So we're using end to end encryption, of course, authentication systems so that only lawyers, of course, would be able to see their client data. And then, we have our proprietary systems that sit within, Casebling and Secure Cloud systems. And any, external API service that we use for AI. We have special agreements so that the data is not retained beyond to the duration of the operation. And some of those special contracts we got actually because we are part of PRA. And so we were able to, have, you know, connections at top AI companies and, you know, had the privilege of being able to talk directly to these, stakeholders and negotiate specific contracts that would help us with, our, PII data.

James Pittman: So, Kaleo, broadly speaking, you know, what what role overall do you see AI playing in shaping the future of immigration law practice? And if you wanna expand to other areas of law and talk in a in a more broad fashion, feel free. And then

Khalil Zlaoui: how how's CasePlank positioned to lead in this regard? So at CasePlank, we don't just take away admin legalese work with AI. We completely redefine the role of immigration lawyers. Empowering them to oversee rather than execute mundane tasks is, how I think about our philosophy. And so really it's taking away all of that manual and processing work and research work, and enabling lawyers to focus on really what's important, which is strategizing for the client case, and then nurturing that relationship with their, with their client. And I think what is happening across, the legal industry is that is AI is definitely transforming the way that lawyers work or will work. And it's estimated that I think 40% of, legal tasks can be automated with generative AI right now in the US. So I think that, really, the way that lawyers work is going to be entirely transformed with AI.

James Pittman: And what sets CaseLink apart from other tools or software that are already available for for immigration case management and and that use AI?

Khalil Zlaoui: Yeah. So we're building meticulously engineered workflows for writing intensive visas. So there's nothing like that right now. I know there are a few startups working on automation in this space, but because this is such, you know, recent technology and because the the workflows are very advanced, it's not easy to automate. And that's what our goal is at CaseBlank, to automate really these writing intensive, complex, workflows. And that's our moat. That's our differentiating factor. And we do this by building very meticulously engineered process end to end to fully automate these complex visas.

James Pittman: Look. Sticking with the end to end theme, you know, when you think about end to end automation of all the processes in an in a law firm, you have to have your technology stack integrated to the greatest degree possible. So does CaseLink integrate with other products, and are you planning any any other integrations, with other lawyer tech tools?

Khalil Zlaoui: Yeah. We've definitely been, thinking about that. So the goal is not to reinvent the wheel. Right? We know that there are great tools out there, that lawyers use. And so we don't want to reinvent what's already out there. We just we really are wanting to build, standalone software that fits where the needs are. And we know that right now, there's the huge demand for automation in in in these complex, writing intensive workflows. So if there are opportunities for integration, we're definitely open to those and we're exploring them.

James Pittman: Alright. So, Kahlil, my understanding from from looking at your website is that you're currently in beta. Are you are you planning on launching at scale in the near future?

Khalil Zlaoui: We are in beta with a cohort of, limited users Because like we we talked about this before, it's very complex automation and we wanna make sure that we are building the right, guardrails and we're iterating with our current users, and, you know, asking them for feedback and learning from them to build really the best product possible. But pretty soon, I would say in the next month, we're gonna release the product to wider audience, and we're looking forward to more sign ups. And so we're very excited about that.

James Pittman: Alright. Excellent. Well, what advice do you have for aspiring innovators looking to make their mark in the legal tech industry? It's a very competitive space right now

Khalil Zlaoui: to be building in, legal AI, but it's such an exciting space. So, I would say if you are aware of a problem and you think you know how to solve it and you're passionate about it, definitely go for it because there's so much to be done right now. And it's such an exciting, time to be building, new things because the technology that we're working with is just unheard of, and just such, a fun time to be building a product.

James Pittman: Julio, it's really been fascinating hearing about CasePlink and and your current stage of development, what you have planned. But how can how can people, you know, get ahold of you, get a demo, sign up for access? Tell us about it.

Khalil Zlaoui: Yeah. Absolutely. So, they can schedule a an onboarding, demo with us, and we would be more than happy to show them the capabilities of Case Planck, and we're looking forward to expanding, our cohort of beta users. And so, anyone who's, who's excited about Case Planck, we're happy more than happy to, to show them a demo.

James Pittman: Again, thank you very much, Khalil Zlaiwei, for joining us, founder of Case Blank. It's it's been a great discussion. It's very exciting, and and it sounds like you have really an amazing an amazing team there, really doing cutting edge things. So I'm really excited to see how it develops over time. Thanks so much for coming on the program.

Khalil Zlaoui: Thank you so much for having me, James.

‍