Company Updates

Akvelon Uses AI to Create Meeting Summarizer Service

Meetings are the most widely used tool for company engagement and colleague interaction. Meetings should be a time for exchanging ideas and participating in deep conversations, but are typically spent with noses buried in notebooks or colleagues frantically typing every word they hear.

A team of our developers wanted to find a way to enhance meeting efficiency for offices around the world. Knowing how stressful note-taking can be, our developers wanted to create a service that would take some of the stress out of meetings and conference calls.

After choosing to participate in Microsoft’s “Innovate AI”, a global artificial intelligence competition, our team set out to create a service that would make note-taking easier while utilizing artificial intelligence to deliver their solution.

From there, our team created the Meeting Summarizer, a note-taking service that uses AI to transcribe conversations and highlight important topics that are discussed in meetings and conference calls. The Meeting Summarizer takes all of the information from meetings or conference calls and compiles it into a short, concise summary that can be sent out to all meeting participants. All the user has to do is add the Meeting Summarizer to a phone call or Skype/Slack call and let the service take over! The Meeting Summarizer is available on demand, day and night, 24/7 with no need for human assistance.

Even the most skilled note-taker can miss a few words or accidentally misquote a presenter. While these mistakes may seem minor, they can lead to confusion and the spread of misinformation. For example, something as simple as writing down the wrong meeting time can cause coworkers to become frustrated with one another, putting a strain on colleague relationships. The Meeting Summarizer takes accurate notes that eliminates human error while using technology such as NumPy, PyDub, pyAudioAnalysis, Sumy, Gensim, and NLTK to ensure quality and consistency. The system also uses 3rd party speech recognition, such as Microsoft Bing Voice Recognition and Google Speech Recognition, and a proprietary summarization engine that is based on an ensemble of models and neural networks.

Akvelon employees smile during morning meeting

The Meeting Summarizer is also very cost effective and can help companies save thousands of dollars a year. On average, a secretary in the United States makes around $33,000 a year (Glassdoor estimate) and spends around ¼ of the year, or approximately 520 hours, taking notes during meetings. That averages out to around $8,250 a year just to have your meetings transcribed.

Akvelon employees brainstorming

The Meeting Summarizer can help any company, large or small, become more efficient and productive by eliminating the need for manual note-taking during meetings. Utilizing the Meeting Summarizer will put the focus back on what’s most important: the conversations had and the connections made during meetings and conferences.

Akvelon colleagues laugh during coffee break

We would like to thank the team responsible for making the Meeting Summarizer a reality! Our Akvelon dream team consisted of Andrey Chekh who was the project manager and thought up the Meeting Summarizer as well as Danylo Kosmin and Kostiantyn Isaienkov who utilized AI and NLP to work on the summarization for this project. Thank you Andrey, Danylo, and Kostiantyn for helping us showcase yet another way that Akvelon uses AI!

Want to learn more about the Meeting Summarizer? Read our case study here.