Having worked previously on teams in both small startups and large corporations, we recognized the many challenges surrounding the improvement of team dynamics, communication, and management. A huge market exists aimed at facilitating workplace productivity, but still the curse of wasteful daily meetings plagues companies around the world. We wanted to take a fresh shot at the problem leveraging the powerful analytical tools available through IBM's AI.
What it does
Synergy is a tool for team managers that is used to record company meetings in the background. After the meeting ends, it sends a Meeting Report to managers that offers them a wide variety of insights into how to improve team productivity. In particular, personality profiles of individual parties are constructed, the tone of the conversation over time is mapped, and an Impact Score is assigned based on how focused team members remain on the key entities and concepts.
How we built it
We utilized a web front end to enable microphone recordings. Feeding that into a Diarization Service splits the meeting audio into separate speakers. From there, the IBM Speech-to-Text API transcribes the meeting and provides text data to three Artificial Intelligence engines. IBM's Tone Recognition Engine highlights specific moments of the discussion that could have been rephrased and suggest improvements for feedback styles. Next, the Personality Engine generates profiles for all parties that enables Synergy to determine their role in the meeting as Facilitator, Analyst, etc. Lastly, the Cognitive Search function allows us to extract relevant entities and concepts throughout the discussion to model overarching meeting themes as well as progress, productivity, and communication efficacy.
Challenges we ran into
Finding and understanding the Diarization Service was a difficult portion of our project. Ensuring quality output from the IBM API given limited input is also a persistent challenge.
Accomplishments that we're proud of
We were especially proud of our design thinking methodology that lead us to solving a highly interesting and relevant problem in a unique way. Grappling with and mastering the relevant technologies was also an incredibly difficult, but rewarding experience as well
What we learned
We discovered that communicating between different platforms, languages, and services was full of pitfalls and errors that we had to face.
What's next for Synergy
There are so many amazing directions to take a product like this. In particular, there are a million sophisticated metrics that could be used to help inform team productivity that could be further developed. Also, integrating with other technologies including email and a web portal for managers.