The third post-hackathon interview: After InteliLex and DoxyChain, it’s time for bSure, the team that took third place in the Polish phase of the Global Legal Hackathon.
Justyna Zandberg-Malec: During the hackathon you worked on an application that points out to freelancers contractual provisions that are disadvantageous to them. Where did you get this idea?
Sabina Łobocka: A colleague who wasn’t taking part in the hackathon suggested it to me (and allowed us to use it). He was signing a contract with a residential real estate developer and didn’t entirely understand all the clauses. It took him a long time to check whether any of the clauses were unfavourable to him. That’s why we thought of an application that ordinary people could use to protect against irregularities and negative legal consequences for them.
During the hackathon, our concept evolved significantly. There were lots of conceptions of the target group we should be addressing. Ultimately, after review of our solutions by mentors, we decided on civil contracts with freelancers hired to perform specific tasks.
Jakub Kubalski: There are contracts that it doesn’t pay for a freelancer to send to a lawyer, given the total value of the contract. Young entrepreneurs often strive to be self-sufficient. The application is designed to help them analyse whether a given clause—for example concerning contractual penalties or noncompetition—is acceptable to them or not, or whether it is worth seeking legal advice in the specific case.
During the hackathon, we were aware of situations where our app would come in very handy. An event producer handed a contract to performers that ran to 5 or 6 pages, which they signed without analysing it thoroughly, knowing only that they would be paid PLN 500. Our app would break down for them in several points what the contract involves and more or less what it contains. The app would display simple messages, written in clear language, but would obviously not take the place of a thorough legal analysis. This is intended to be a tool making it easier to take business decisions.
But I understand that in that situation the contracts were on paper, not in electronic form. How do you feed the content from the hard copy into the application?
Jakub Kubalski: That’s not hard. There are tech solutions, such as OCR, that extract text from a picture or graphic file. There is a risk of errors, obviously, but the application could for example display a message that the picture should be taken in better light or with a flash, so that the text is more visible. But that’s just technical stuff.
Kamila Klimowicz: We started from the assumption that the app must handle images of contracts, because if someone wants to make a decision on the run, they probably have a hard copy of the contract in hand, not an editable version.
Sabina Łobocka: But of course we don’t rule out text files either.
I understand that an app like that would have to have at its disposal a huge database of clauses which it could compare against the contract?
Jakub Kubalski: Not necessarily huge. We drew up a working list of keywords that the application will catch and assign a message, symbol or colour to them. For the purpose of the hackathon, we decided that colour would be the easiest. Green, yellow or red could signify the importance of the particular provision. Green would be reserved for basic issues like identification of the parties or their tasks, and red for example for a contractual penalty or ban on competition. We must remember that the target is small businesses concluding contracts for small values, which mostly aren’t going to be reviewed by a lawyer.
So it’s a signal that says, “Pay attention to this, here’s a clause that could be dangerous.”
Jakub Kubalski: A clause that would be worth consulting with a lawyer—it could be a friend. We assume that anyone who is in business probably knows a lawyer. Certainly their accountant would know somebody like that. So the report generated by our application would enable a decision on whether to sign the contract now or consult it with a lawyer. Based on our observations, freelancers usually don’t read contracts, but at the same time they feel uncomfortable signing something they don’t understand.
Kamila Klimowicz: In other words, in the first step we don’t have to assign so much authority to the app that it would say that something is advantageous or not. It should only alert the user to potentially risky elements by assigning the relevant message to them. Let’s suppose that provisions concerning contractual penalties are marked in red. The message could be conveyed alongside this, “Check that the penalty is proportional to the value of the contract.” So the freelancer could check this, but would have a simple indicator of what to pay attention to. For example, if a penalty available to us is limited to the situation where the other side does not pay, we would know that realistically it would not be applicable.
What about liability? Suppose the whole contract flashes green, and the freelancer goes ahead and signs it, but later a dispute arises where it turns out that the contract contains an unfavourable clause.
Jakub Kubalski: Our aim is not to take the place of legal analysis, and thus we obviously can’t assume liability in this respect, and this would be clearly stated in the terms and conditions for the service.
Sabina Łobocka: We’re more inclined to point out things that require analysis, not necessarily confirming that the contract is OK.
What business model do you plan? It seems you’re targeting your product to the weaker party to the contract, i.e. the one holding less funds.
Jakub Kubalski: This issue is still the subject of discussion. Initially we didn’t want to impose any barrier in the form of payment. We are considering a free model with monetisation in the form of ads, although pay models without ads could also be considered, more advanced with additional paid modules. They could introduce the beginnings of a legal analysis, suggesting a solution to the lawyer examining the clause.
Sabina Łobocka: Verification of the contracting party could also be free: whether the party is reliable, how long they have been operating on the market, and whether there are any indications that it could be risky to cooperate with them. Then the pointers concerning the contract itself would be paid.
What’s the most difficult at the moment? The concept seems brilliant in its simplicity, but now how will you implement it?
Sabina Łobocka: We don’t have with us at the moment our two programmers, Marcin Dahlen and Marcin Wylot, who could discuss the technical details. But we know from them that they planned to use neural networks, and specifically Bi-Directional LSTM with Attention Flow. On the substantive side, we must provide some sample contracts and keywords, and also verify that the algorithms respond appropriately to what they are fed. The most work would thus be required on the typical database side. We don’t anticipate major problems on the frontend and UX side.
Jakub Kubalski: With a developed methodology for creating the database, it will be pretty easy to implement the application in further markets. And scalability of a business model is essential today. With the developed methodology, we could commission supplementation of the database with relevant keywords in other languages. After all, problems of this type aren’t encountered only by Polish entrepreneurs.
Sabina Łobocka: We can also easily expand the application to other types of contracts.
Jakub Kubalski: A general growth area is legal language. Another popular topic recently has been legal design, i.e. conveying legal content not only in words, but using ideograms, drawings and schemata, or through appropriate layout of the content of contracts so they are more understandable. I imagine that an application could propose a contract that is already trimmed and more user-friendly. If often happens that contracts are intentionally written opaquely. And the final provision on page 40, printed in 8-point type, might concern the most important issue, such as automatic extension of the contract. I admit that once I was taken in by such a scheme myself.
Sabina Łobocka: When we were selecting the target group during the hackathon, we considered for example contracts with telecoms or residential developers, but there we would have little room to manoeuvre. We could raise users’ awareness of what they’re signing, but the chances of negotiating such contracts are small. You either sign or you don’t.
Kamila Klimowicz: Later we narrowed the target group to freelancers, mainly in the IT field, because it is a large and rapidly growing number of people who are young and inexperienced. On one hand they often ignore contracts, because they don’t seem important to them, but on the other hand they are not in a position to decipher them. Our application could raise the safety and legal awareness of such people a bit.
Jakub Kubalski: For example on the issue of copyright. Transfer of economic copyright means that people can’t use what they themselves created. This is an essential issue particularly on the IT market, where solutions are often replicated. And during due diligence I often see contracts where an IT firm gives away its core business every time it enters into a contract with a customer. One of the aims of our app is to flag such items as material issues.
When can we expect to see it in Google Play?
Jakub Kubalski: Let’s say in mid-2020.
Sabina Łobocka: To be honest, each one of us is working every day on something else, and it’s hard to find time to develop something together, to research needs, seek sponsors, and verify the target group or business model. We need incentives, and certainly this interview is one of them.
Let’s talk a bit about the hackathon. Is that where you all met?
Jakub Kubalski: Yes.
Sabina Łobocka: I came with a preliminary concept, but I knew that it would be significantly modified. But I really wanted to create something for people. I have no idea how lawyers operate, I’m just learning that, so I can only speak as a layperson. I managed to recruit the group. It started with Jakub, then we were joined by Marcin, who is a developer and also has graphic skills, and once you have a developer onboard that is your first success. Our lawyers, who have great experience in civil contracts and the media industry, were a great fit and worked well through that period. Certain topics had to be cut short, and we had to work together and adopt an action plan. The mentors were very helpful in that, constantly challenging us, asking questions about what it was supposed to look like, what it would do, what for and why.
Kamila Klimowicz: All of these uncomfortable questions ultimately led us onto the right path.
Sabina Łobocka: Although it was a little irritating. We’re working feverishly, surrounded by sheets of paper and felt-tip markers, and suddenly somebody walks up for the nth time and asks, “What are you doing here?” But it proved helpful and necessary, because indeed our visions were wandering off in all directions.
But are you pleased with the direction where you were aimed, or was something lost?
Jakub Kubalski: We can’t say there is a loss when we were gaining experience. For me this was the greatest value of the Global Legal Hackathon. I was participating for the second time. The first time we had a bit different project, a kind of chatbot. It was fascinating for me to work with IT people, because as lawyers we don’t get this opportunity. Ever. We don’t know what the process of building a database looks like, creating a nexus, coding functions. For me it was a great experience. And for me, that is what makes this event important.
Should lawyers work more closely with programmers?
Bartosz Prorok: I heard an interesting statement on a podcast once that new technologies grow exponentially and law grows linearly. I completely agree. Maintaining the current approach to the development of law, at some moment a situation will be reached where new technologies and programmers are so far ahead with their solutions that cooperation between programmers and lawyers is unavoidable. That should be viewed as a positive development. That is our future, as IT is developing so far and so many interesting solutions are being created.
Jakub Kubalski: Some claim that in the future there won’t be any lawyers, only legal engineers capable of operating research and analysis systems. Humans will be the face, not the engine, of legal analysis. This is the direction where the growth of AI, smart contracts and blockchain is heading. I’m talking here about a horizon of a couple of decades into the future.
Bartosz Prorok: But in the shorter term, this application means that lawyers will have more interesting work. We often encounter relatively simple cases, where it takes more time to analyse specific elements and documents than to develop responses. Such routine matters could be handled by an application, and the lawyer could focus on strictly substantive issues. I’m looking forward to an application like that to relieve me of some of the paperwork.
Sabina Łobocka: We very much recommend participating in the Global Legal Hackathon. The encounter between lawyers, programmers and businesspeople is an interesting experience.
Is there anything there that would be of interest to people who are neither lawyers nor programmers?
Sabina Łobocka: I actually work as a project manager, so I very much appreciated all the tools that help convey that the concept alone is not everything. It’s essential to know what problems our product will solve, what are its advantages, whom it is targeted to, and what the sales channels are. Now everyone wants to establish startups, but this requires a pragmatic approach, to see how many questions must be answered before the ultimate business can take off.
Kamila Klimowicz: That’s why there’s such a need for people who handle projects every day, not necessarily in the legal or IT industry. They help coordinate the thinking of these teams. Lawyers focus on legal issues, and programmers on technology. It takes someone to bridge these two groups, because they may not be on the same page.
Bartosz Prorok: People from practically any other sector are needed. The aim of the hackathon after all is to find solutions, and a person unconnected with IT or law may come with a problem to be solved. Moreover, you don’t have to be a lawyer or IT person to think of such a solution. Anyone can contribute something that moves the project forward. So while it is supposedly a legal and IT event, anyone could try it at least once.
Interview conducted by Justyna Zandberg-Malec