Product builders around the world frequently trade-off between three strategies
When should you (re)-build your models and invest in iteration to get great quality? When should you borrow high-performance APIs and free your developers to make the next big thing? And when should you buy solutions for your problem?
What is VideoKen?
VideoKen is an AI-powered video interactivity solution that transforms learning videos into interactive and immersive experiences. In this age, attention is at a premium, and learners have access to more information than they can consume. “VideoKen’s purpose is to take learner engagement to the next level by making it easy for people to discover and interact with videos,” says Vishnu. VideoKen offers automatic video chaptering, in-video quizzes, deep video search, video analytics, closed captioning, video hosting and other features to make this interaction rich and impactful.
What happens behind the scenes at VideoKen?
VideoKen’s solution uses AI to track, predict and improve learner engagement using in-video interactivity and actionable insights. The goal is to minimize the problem of poor engagement among learners.
Over 100+ learning leaders leverage VideoKen to help learners engage and interact with videos using an auto-generated navigable storyline, deep video search and in-video assessments. These leaders also tap into VideoKen’s in-depth analytics to gain insights into their learners’ preferences and behavior, which is used to optimize and prioritize their learning content.
“VideoKen’s customers have been successful by driving 3x more engagement and retention on their learning videos,” according to Vishnu.
VideoKen’s development journey
VideoKen’s journey started by developing in-house technology (now patented) for improving engagement. On top of their five patents, VideoKen needed to be able to offer a comprehensive solution to meet growing customer needs. “The engineers were eager to build in-house models, but it turned out we couldn’t do it all by ourselves,” says Vishnu.
Moreover, VideoKen needed an adequate cloud computing infrastructure which led to exploring and experimenting with various options before settling on Azure.
Developing in-house machine learning models for speech-to-text transcription
Jeril Sebastian, Head of Engineering at VideoKen, shared insights on using some of the applied AI services. “Most of our services rely on VideoKen’s own AI technology. One thing that we do rely on externally, is Azure’s Cognitive Services for transcribing videos.”
Jeril explained that before leveraging Microsoft Azure and its Cognitive Services, VideoKen was using a combination of its proprietary speech engine along with some services of another cloud provider. “It was taking up a lot of time and wasn’t feasible for us to pursue this approach long-term,” Jeril said, adding that the company was spending twice as much money on the in-house model compared to Azure’s speech API. “Even though everything was working well with our AI speech-to-text solution, we needed to focus on our key strengths and have our developers work on solutions that are important to our customers.”
After discovering that hosting GPU instances and training models in-house was taking up too many resources, VideoKen decided to test different options, with Azure’s speech to text proving a perfect fit. “Speech to text is a single building block of a much bigger infrastructure, so we couldn’t afford to lose a lot of time and energy, despite being satisfied with the overall performance of our home-grown technology. Therefore, replacing it with Azure was the most logical step for us.”
VideoKen’s journey post-Applied AI
Video transcription has become a norm nowadays. One of the most important aspects of making videos more immersive is ensuring that learners do not miss any important information. Having realized Azure Cognitive Services’ power, Jeril and the engineering team managed to cut the cost of video transcription by 50%.
“We used to have an ML engineer just to train the model, improve it, run the GPU instances, and so on,” Jeril said. “This was a full-time job for one person, and now it takes just a few hours a week for me to take care of that entire pipeline. Our ML engineer now has more time to focus on researching and upgrading other services offered by VideoKen, instead of doing the grunt work to keep our in-house transcription model competitive.”
Processing thousands of videos with Azure
Apart from Cognitive Services, VideoKen also relies on other Azure benefits. “We use two specific pieces of technology. One is called Azure Functions, and the other is Azure Container Instances,” explained Jeril. He added that the two tech pieces allowed VideoKen to scale up and down elastically, “We get thousands of videos at a time from our customers, so quick scalability is essential.”
By relying on Azure Container Instances and Azure Functions, VideoKen has managed to set up a pipeline which allowed the company to scale. “When there are no videos to process (theoretically), it costs us nothing,” Jeril said. “On the other hand, when there are thousands of videos coming in, VideoKen scales up automatically and is able to handle that load.” At full capacity, VideoKen can process approximately a thousand hours of video per day.
Jeril shared his excitement with Functions and Container Instances. “I really enjoy working with Functions, as it has saved us a lot of time. This is a realm of technology called serverless, and it has made management and maintenance really easy.” Moreover, Jeril stated that Container Instances helped VideoKen’s scalability in a unique way, given that such technology doesn’t exist on other cloud services.
The future looks bright for VideoKen’s partnership with Microsoft, “Right now, we’re happy with Azure, and the plan is to stay with the service for a long time,” said Jeril.
He expressed his satisfaction with Azure’s transparency regarding costs. “They are trying their best to show exactly how and where our money is spent for Azure’s services. Whenever I visit the cost management section, I can see exactly where we’re spending money and how much is being spent.” The main advantage of this, according to VideoKen’s Head of Engineering, is that the company can make better decisions on how to use the Azure budget.
To get started with Microsoft for Startups Founders Hub, sign up here.