Industry: Artificial Intelligence/Sales & Marketing
Founding Student Name(s): Ankush Bikkasani, senior pursuing Bachelor of Business Administration degree, Indiana University Kelley School of Business
Brief Description of Solution: Video production is an incredibly time- and cost-intensive process. Most companies spend an average of $500-$1,000 and 4 hours per minute of the finished video. Last year 95% of companies said they wanted to produce more video, but were limited by time and budgetary constraints. Ultimately, this means that video production isn’t scalable.
Deep Word allows users to create videos of people talking, without having to repeatedly film them. Users select one of our video actors or upload their own (a short video of the person they want talking). Then they type the text or upload the audio they want this person to say. Within minutes, they have produced a video of this actor speaking their desired audio. Here is an example. Our website will work with any language, allowing one to produce the same piece of content for a multitude of languages. Updating a video can be as easy as changing its script.
We also have developed an API that will allow for integration with a range of other services. Instead of having to create videos one by one via our website, our API and its integrations can be leveraged to automatically personalize videos for individual customers and employees at scale for marketing, sales or training purposes. Companies essentially can connect our API to their databases, CRM’s, or LMS’s in order to generate as much personalized video content as they need.
What led you to launch this venture? My background lies in professional video production. I’ve been freelancing it at a high level for the past 10 or so years. Overall, it’s an incredibly repetitive process involving the same people talking about very similar things over and over again. I began looking for ways to automate this process in the summer of 2020. However, I wasn’t able to find any solutions to this with any existing software solutions. From here, I contacted some of my data science friends at IU’s Luddy School of Informatics, Computing and Engineering and we began experimenting, developing, and testing different machine learning solutions for this. Eventually, we settled on the prototype currently deployed to our website and have iterated on it significantly since then.
What has been your biggest accomplishment so far with venture? To date, we’ve raised ~$175,000 from various competitions and funds. This includes being marked as winners at Elevate’s Pre-Seed & Seed Competitions, as well as the Cheng Wu Innovation Challenge and the Indiana University Kelley School of Business’ Clapp Idea Competition.
Since launching our website in November of 2020, we’ve scaled to just over 31,000 users, of which are currently generating over 4,300 videos per week.
As we have started to roll out early versions of our scaled video personalization solutions via our API, CRM, and Zapier integrations, we’ve also begun to attract and initiate pilot level agreements with several sales, marketing, and web development firms over the next several months.
How has your business-related major helped you further this startup venture? Perhaps the most useful outcome of my business major has been the connections I’ve made through it. Countless students and organizations I’ve come across have gone out of their way to leverage their skills and networks to help and guide our startup when we were first starting out. Students and professors alike have introduced me to countless advisors, mentors, similar startups, and investors over the last 12 months alone. I know that myself, our team, and startup will continue to benefit from these connections and partnerships for many years to come.
Which business class has been most valuable in building your startup and what was the biggest lesson you gained from it? Easily the most useful class I’ve taken has been K201. Excel is a tool I use almost daily for our accounting, budgeting, resourcing planning, and future financial modeling. Overall, the class has taught me the sheer usefulness of effective planning, especially as projects began to increase in complexity.
What business professor made a significant contribution to your plans and why? I’d like to highlight two people.
A good friend of mine took a high-level entrepreneurial class with Professor Regan Stevenson and strongly recommended that I seek him out for guidance. At that time, we were early with Deep Word, and still experimenting in the ideation phase. Regan gave me great advice around building and testing prototypes for our early software and even introduced me to some computer vision Luddy professors who helped us through our early product architecture development. He also gave me some incredibly actionable feedback around attracting both early customers and investors.
Another friend at Kelley also introduced me to Cy Megnin, an adjunct entrepreneurship professor at IU’s campus in nearby Columbus and the entrepreneur-in-residence for Elevate Ventures, a development organization that nurtures and develops emerging and existing high-potential businesses in Indiana. Cy has had extensive experience as a founder in tech startups, including a number of very successful exits, and has been closely advising our startup since its inception. I credit Cy for a lot of our prep work for investor and competition pitches that we received funding from. He’s also been incredibly resourceful in connecting us with other entrepreneurs, startups, and established companies interested in potential partnerships. Overall, Cy has been an invaluable resource and great friend of mine since I met him last summer.
I can definitely say that we would not have gotten nearly this far without both Regan and Cy’s guidance and support.
What is your long-term goal with your startup? Ultimately, we are developing video-based solutions for companies to connect with their customers and employees on a personal level. We believe that scaled & personalized video production, something which has only been made possible by recent developments in machine learning and computer vision, are the best ways forward in doing this. Our goal is to seek acquisition after the next 3-5 years.