Live Brain
Artificial Intelligence solution embedded into your Live Stream Shopping sessions
Regardless of the software solution that we choose to deploy, here at Oveit, Inc., we are guided by some major beliefs.
One of them is that technology should always make our lives better and our work easier. This was and still is, what unites us as a team and makes us go further in the hardest days. Whenever possible, wherever possible, technology should take care of repetitive tasks and humans should focus on creating meaningful experiences.
Another major belief that makes us roll our sleeves and get to work is this one: small communities, united by a common goal, are what make us better as a whole. Over the last few years, we saw that ideas have the power to gather people around them and move things around for the better. Businesses, brands, non-profits, individuals, regardless of where the idea comes from, the end result is often the same: people unite behind them, create communities, and challenge the status quo. And this is absolutely magnificent.
We have used this short introduction to highlight the fact that we don’t create software, we create solutions. We don’t create technical solutions for the sake of technology. And each feature that we work on has its own role in the bigger picture.
Live Brain, our AI power solution embedded into Streams.live, makes no difference. We’ve built it to make your life easier. We’ve built it not only to take care of the repetitive tasks but to also ensure the integrity of your brand and community. Here is what we mean by this:
Hate Speech and Profanity Detection
We don’t tolerate hate. Neither should you. A live commerce session should be fun and engaging, it should create a feeling of trust and belonging. But how could you possibly detect hate and profanity when thousands of viewers are using the chat option to engage with you and other followers?
Using AI, Live Brain takes this problem off your shoulders, allowing you to focus on creating memorable live stream shopping sessions. We won’t go very deep into the technicalities but here is some extra information on why you should leave these tasks on our system.
We’ve chosen to use the most performant method of pretraining language representations to make sure that we understand and block complex forms of hate speech, not just words that may seem inappropriate. It’s a neural network that was pre-trained on data sets provided by Twitter and Wikipedia, thus helping it get accommodate with all forms of hate speech. And it’s using the Tensor Processing Unit, allowing it to efficiently process big volumes of data.
Sentiment Analysis
Sentiment analysis is important because it helps you better understand the overall opinion, attitude, and sentiment of people that take part in an experience. And from a business perspective, this is extremely important: you want to know what viewers feel about your work, right?
When having thousands of messages it may become difficult to see the big picture. What are all these people talking about? Are they happy? Or, on the contrary, this whole experience left them with a negative feeling? It’s important to understand how people interact with your content. And again, this is a task that machines do a lot better than humans. Using the models that extract high-quality data from the text inputs (chat and Q&A), we will tell you if your live shopping session had positive or negative feedback from your viewers. And trust us, it does it faster and better than any human being as a dataset collected by Stanford University of over 84 million text entries has been used to train the model.
What’s next
But Live Brain can do so much more than just this. Our next steps are into summarization. We are working on “teaching” our virtual colleague to help you with something quite revolutionary: if you or your colleagues engage in one-on-one conversations, our solution will be there to “take notes”. And at the end of the meeting, it will help you with a brief summary of the discussion. A Conversational Summarization model that pinpoints the most important ideas from the conversation.
Why? Because we feel that during such a meeting, your partner should benefit from your full attention. You will have your own relentless colleague that will “take notes” and help you with a concise summary pinpointing the main ideas of the discussion. But more on this will follow.
