Could an algorithm make divorce and custody disputes simpler and less stressful?

Global Legal Hackathon finalists explain how their ‘consensus-finding engine’ helps parties resolve disputes — mathematically

Resolving disputes, particularly divorce disputes, can be a timely and expensive process. Custody battles are often fraught with hostility, and financial settlements are tense at best. But imagine if there were an algorithm that could help both parties arrive at an agreement that worked for each of them without the added cost or time inherent to legal proceedings.

A group of legal and tech entrepreneurs in Berlin may have found that exact solution. A “consensus-finding engine,” as its creators call it, Sondier.ai is the product of deep industry knowledge, inspiration and relentless work that began during the first round of the Global Legal Hackathon in Berlin, hosted by hacking.law.

In just a couple short days, the Berlin hackathon team will compete against 13 other teams at the final round of the Global Legal Hackathon in New York. (It’s not too late to support the team by joining them on April 21 in New York ! Event details here.)

The team consists of Dr. Konstantin Bertram (LAWLIFT GmbH), Johannes Wagner (LAWLIFT GmbH), Luis Hoefer (superlumen.io), Dr. David Linner (co-founder Smartlaw), Dr. Ralf-Michael Schmidt (co-founder Smartlaw), and Stefan Hoehn (Metamorph GmbH).

We spoke to the team about how their ‘consensus-finding engine’ helps people solve disputes and how they plan to expand on it.

How did you form your hackathon team?

Ralf-Michael Schmidt: David, Stefan and I worked for Smartlaw.de, an online legal document service, co-founded by Ralf in 2012, offering customized legal documents for B2B and B2C target groups. We met Luis, Konstantin and Johannes at several legal tech events. Konstantin is co-founder of Lawlift.de, a document automation solution for lawyers and legal departments, for which Johannes also works. Luis co-founded superlumen.io, a tech firm specialising in rapid prototyping.

Why did you decide to build a ‘consensus-finding engine’?

Ralf: Negotiating an agreement or resolving a dispute, such as a divorce dispute, is often very inefficient, time-consuming and expensive. But it is possible to make it efficient, easy and hassle-free for all parties.

Can you describe how Sondier.ai works?

Ralf: Parties using Sondier.ai start by answering a questionnaire to define their positions on disputable issues. At the end of this questionnaire the parties can rank their positions. Ranking means that each party can determine which of their positions are most important for them. Of course all this happens totally separate for both parties.

Our tool helps both parties to achieve a compromise solution within a few minutes. An unbiased algorithm helps parties to create a consensus basecamp. In cases where remaining positions are too far apart, a mediator or arbitrator is suggested.

What was the biggest challenge in developing — and launching — the platform and how did you overcome it?

Johannes Wagner: One of the biggest challenges was to decide on a topic on the basis of which we could explain our tool in the best way. We found the topics of negotiating term sheets (M&A) or finding solutions for licensing infringements particularly exciting, but ultimately decided on the topic of divorce, which is easy to understand for anyone, especially non-lawyers.

You mention Sondier.ai’s utility in solving divorce disputes. Does it currently support other types of dispute resolution cases, and do you see it evolving beyond family law applications?

Johannes: So far we have focused on divorce for exemplary purposes and users can already answer some basic questions, but we are still improving this and are coordinating with experienced mediators.

The topic of divorce is only one example of the numerous possibilities of our tool. Sondier.ai makes it possible to resolve conflicts on different levels from neighboring or investment disputes to negotiations to form a government or international conflicts. Third party authors as well as ourselves can create consensus finding templates by simply adding a new questionnaire. It is therefore our vision to facilitate conflict resolution in various areas.

Is there anything else on the market like this?

Johannes: According to our research, there is no comparable tool on the market. Our solution is characterized by the fact that the answers are not disclosed directly to the other party, which speeds up and simplifies the entire process. Only during the conflict resolution process the approximate positions can be identified. Furthermore, our intelligent conflict resolution proposals (“micro-consensus”) make it possible to find balanced solutions in a simple and fast way.

If there are still open questions at the end our partner mediators or lawyers will resolve them in a targeted and focused manner at low cost because of reducing the scope of the dispute.

Where does AI come into play?

David Linner: We plan to apply machine learning on the behavioral characteristics of parties in negotiation scenarios. This is supposed to improve the micro-consensuses proposed by the algorithm and shorten the negotiation process.

When would lawyers and/or mediators come into play, if at all?

David: Breaking a tie in a negotiation may eventually require an understanding of the parties’ inner motivation. At this point fairness is no longer a “mathematical matter”. Finding a consensus may require the empathy of a human.

What do you think will be the biggest challenge going forward?

David: Developing business around the service. The legal tech market in Germany is still quite young. Convincing potential customers to give a solution like sondier.ai a chance will be hard work. We are currently discussing how to enter the market and to address this challenge.

To watch the Global Legal Hackathon finalists compete in New York on April 21, attend the GLH Gala (event info here)! And to get the latest updates on Sondier.ai, follow them on Twitter (@sondier_ai ).