Anna the AI Law Assistant was inspired by the need identified in Castrén & Snellman - Accessing the Law. In the challenge, Castrén & Snellman identified as a goal to improve information flow between clients and lawyers and to bring innovation into legal services.

We decided to create a chat bot which would help lawyers to collect standard information and documentation from new clients through a conversation with the bot. The bot would then forward the collected information to the lawyers for processing.

What it does

Anna the AI Law Assistant helps lawyers to collect standard information and documentation from new clients through having a conversation with the bot. The bot then will forward the collected information to the lawyers for processing. This brings a new way for clients to reach out to legal services at any time of the day, which is more intuitive and easier to use. It also makes lawyers' work more efficient by providing them a tool to take care of the routine information collection.

Anna can also speak and applies voice recognition technology which allows it to understand spoken languages, when clients may not be able to type (e.g. driving in a car), or where clients may have disabilities which prevent them from doing so. Anna is there to answer clients' questions whenever and wherever they are. This makes the law and legal services literally more accessible.

How we built it

We created a neural network in C # and Then we trained its language models on Azure virtual computers. Then they wrote the server part and placed it on Azure. We created a mobile application for Android on Kotlin, which implements the conceived functionality. In addition, we created a telegram bot.

Challenges we ran into

We ran into design problems =) we argued for a long time how to make our application more beautiful. We also faced the problem of speech recognition and the problem of understanding complex contexts. But thanks to coffee we successfully overcome them!

Accomplishments that we're proud of

We are very proud of how the bot understands speech. It turned out pretty clever. We are also proud of our design and the fact that our application helps people solve their problems related to the law. The law is a very complex system and our application will help to optimize this system and make the world better!

What we learned

Developers: Firstly, we learned how to handle complex English sentences and intentions. Before that, we worked only with the Russian language and it was something new for us. We improved our programming skills, creating design and raising servers) And most importantly, we improved our skills in teamwork and international communication. These were unforgettable days!

Non-developers: We learned about the logic how chat bots work and can be made. We also learned about working in an international team. Thanks to the amazing developers we got to work with and learn from, we saw that there are so many possibilities how legal services can be improved with technology. We also expanded our legal knowledge and understanding legal services' IT systems.

What's next for Law Assistant

The possibilities for the AI Law Assistant Anna are endless. Its conversations can be expanded to various fields of interest for Castrén & Snellman to cover many different cases, wherever it is possible to standardize a part of the conversation and information collection from the client.

Anna also learns from each conversation, which makes its output and service better with each conversation.

We also have plans to connect Anna with the data from the Castrén & Snellman CRM to provide information for existing clients about the progress of their project, to ask questions about their invoices, to find information about the correct contact person etc. It could also have an interface which would visualize some of the key data so that it would be accessible for existing clients at any time, anywhere.

Built With

  • azure
  • c#
  • kotlin
  • microsoft-cognitive-services
  • yandex.neural
Share this project: