Watching video news on a mobile device can be frustrating. Clips are short, usually 1 to 2 minutes, leaving viewers to continually search and refresh for each topic of interest. Editorji aims to improve the mobile news experience by allowing users to continuously stream news clips tailored to their interest. The Delhi-based startup wants to become the Spotify of video news, using artificial intelligence (AI) and machine learning (ML) to learn users’ tastes based on factors such as how long they watch a video or what tags they most frequently watch.
The beta version of the Editorji app was released in October 2018, with the full version going live the following March. More than 500,000 people downloaded the app during its beta phase, and its founders anticipate another 2 million users in the first few months of operations. Most viewers are based in India, but Editorji has a growing client base outside the country, primarily Indian expatriates in the United States, Canada, Singapore, and the Middle East.
Editorji prides itself on its ability to curate content and weed out “fake news” often distributed on social media. The business is founded by Vikram Chandra, former Group CEO of NDTV Limited, and a seasoned news reporter and anchor with over 25 years of experience in the television industry. The Editorji newsroom boasts some of the industry’s leading news producers in India. Apart from creating their own stories, the team carefully screens independent publishing partners to ensure stories’ credibility. Most of the app’s content is sourced from established news wires.
The Editorji operation is structured around outsourcing partners dedicated to particular tasks, so opting for a cloud provider to manage its IT infrastructure made sense. The founding team at Editorji and its tech partners at Opalina Technologies had experience working with Amazon Web Services (AWS). They began consulting with AWS representatives upon launching to build a fault-tolerant cloud architecture. “We strongly feel that when it comes to cloud computing, AWS is miles ahead of competitors,” says Biswajit Borkataky, head of operations and platform relationships at Editorji.
The first six months of operations were spent gathering data on usage patterns to start building its ML algorithm. Editorji works with DataMonk, a data intelligence company based in India, to develop its AI projects. At present, DataMonk and Editorji are consulting with AWS teams to implement pending AI tools from AWS such as a new recommendations engine in the testing stage. “We’re hoping to be early adopters of this engine because we’re very pleased with the AWS services we are using so far,” says Borkataky.
Among the various AWS solutions in use, Elastic Load Balancing is one of the most critical for Editorji’s business. News is unpredictable, and a major event could cause traffic on the app to surge by 300 percent, for example. With Elastic Load Balancing, Editorji’s infrastructure scales rapidly to accommodate such spikes. The startup uses a serverless model to scale on demand and keep costs low on slow news days. It is also starting to integrate the app with other over-the-top (OTT) content providers’ streaming media content such as Airtel TV App, which makes it nearly impossible to predict the number of viewers per day.
Editorji relies on Amazon Polly to voice its graphics-based videos and Amazon Comprehend for entity extraction and tagging in the cloud. “We use Amazon Comprehend to perform entity extraction from video scripts. It pulls out all the tags, and the tagging is done automatically. Every time our news editors release a script they don’t have to tag the story again—the extraction picks it up and just does it,” Borkataky explains. Editorji also uses Amazon Elastic Transcoder to convert media files from their source format into versions that will play back on mobile devices.
Editorji subscribes to AWS Enterprise Support as the founders value its 24/7 operations model. Editorji’s AWS representative visits the office at least once a month to evaluate the company’s cloud architecture and make suggestions on ways to reduce costs or increase speed-to-market—both make-or-break factors for a new business.
“AWS keeps capital expenditure down in many ways, so we don’t have to invest in servers and a large number of resources to manage our infrastructure,” Borkataky says. Just one employee is responsible for the company’s cloud infrastructure, and, in typical startup fashion, that person is also tasked with multiple other roles. While the app was in development, the infrastructure manager received support by attending a week-long AWS training session.
Editorji hopes to quickly earn a reputation as the top trusted video news site in the country. Audio and text versions have been integrated with the latest version of the app, and new AI/ML features will be rapidly incorporated this year. Discussions are ongoing with AWS to discover new ways to capitalize on plug-and-play tools on the platform, including the potential for automated translation to new languages.
Editorji is prioritizing the local market in India for now but retains a long-term vision for global expansion. “With AWS we don’t have to replicate our entire setup when we expand into a new market. We have one central place where everything is done and monitored, but can be personalized for each market,” Borkataky says. “Because our architecture scales instantly, it has allowed us to expand our reach in India.”
To Learn more, visit https://aws.amazon.com/media/.