Our inspiration came from our own personal needs to be more efficient with our business tasks related to learning and marketing material production. We needed to add professional voices to campaign and learning videos, but that was an expensive process to go through studios and hire professionals, with lots of back and forth, without always getting the expected outcomes from it. We truly believed we could achieve more by leveraging AI to help us with that. However, when we started looking for solutions for that, we found either deep technical solutions or standalone application, and in order to be productive and maximize our results we needed AI solutions that were simple to use, that could adapt to our workflows and not require us to adapt to the app. As our main business tool is Teams, we thought that the perfect solution we needed was one that would bring AI to Teams, to solve our business problem at the same time that would allow us to collaborate in the same way we were already doing for everything else. Because we couldn't find such tool, we felt that we had an opportunity to innovate and create our one perfect tool. So, we did it :).
What it does
BitFractal help business users, across all industries, to be more productive and efficient by leveraging advanced Cognitive AI capabilities in an intuitive and simple way, allowing users to naturally add artificial intelligence to their daily activities, while collaborating with colleagues and connecting the outcome of the app to existing business processes at easy.
With BitFractal, business users can quickly turn text scripts into audio segments to be used as video voice overs, audio prompts, corporate messages, audio books and more.
With hundreds of voices in dozens of languages, a BitFractal user can quickly localize any audio resource from a translated text input.
BitFractal was designed to run on Teams, and benefits from all its capabilities to add collaboration and productivity for the entire organization.
How we built it
The app was built following all best practices for a Teams App, the front-end is a Typescript React App, using Fluent UI, built exclusively for Teams. See our attached Architectural Diagram for more details.
Challenges we ran into
Building new apps on a new platform with modern frameworks required a lot of learning, and the short timeframe we had for that was the most challenging part. Huge dedication, extra hours of reading and working, lots of testing, made us get to this point. In terms of roadblocks with the technology we didn't really face any challenges, as we learned more about a give technology we were supposed to use in our project, we adapted, and we were able to achieve 100% of our goals.
Accomplishments that we're proud of
We're really proud of how much we were able to accomplish in such a short amount of time. We believe our app is quite complete, leverages many Teams capabilities in it, with advanced features that can really impact customers and that our modern architecture makes our app ready for the future.
What we learned
We learned so much during this project, not only how to build modern applications for Teams, but how ideas can be expanded as we go and learn more about what each Teams features and capabilities can do. We started with a simple concept of hosting the app as a personal app on Teams, with some chat and bot features. But as we developed the project we learned and realized how we could bring way more value to our users from deeply integrating our app with Teams to have our capabilities meeting customers where they are and participating on their business workflows and daily tasks in a way no other app, with similar scope to ours, can do today.
What's next for BitFractal
Next, we're going to enhance our first microapp, Text-to-Speech, capabilities to drive even more value to our users and have our app from an MVP to a commercial state. From there, we'll be adding, over time, new microapps to our product that will work as Lego blocks for users, for example, an user with the Text-to-Speech app, once we make the AI Translator App available, will be able to use that as a standalone feature or use it to translate their new or existing Text-to-Script projects to multiple languages and quickly have their voice overs in as many languages as needed. In the same way, once the Voice-to-text application becomes available, the business user, instead of typing the script for their voice over projects, they'll be able to simple narrate their videos with their voices, our app will capture and transcribe that and from there the user will be able to apply one of our neural voices to build a professional voice over project in multiple languages. And for each and every new feature, we'll be incorporating them to Teams collaborative framework, allowing users to work together on projects, simplify tasks through our embedded bot, leverage message extensions to move information from chat to their projects and backs, and etc. In addition, we plan to expose some of our services through Azure API Management to create a Power Platform connector and allow users to integrate our capabilities to their workflows.