In this edition of Industry Spotlight, let’s take a look at Vadoo, an early-stage video startup aimed at disrupting video streaming using P2P and AI.
Video Streaming Can Get Expensive
Video streaming is both expensive and complicated, and the industry has done a lot to mitigate both problems.
Streaming costs and user experience can be improved by encoding more efficiently (newer codecs, encoding at resolutions + bitrates suited to your target audience’s geography and network conditions), using CDNs with more edge servers, predictive caching, per-device manifests, and so on.
Despite all this, companies had to limit their streaming to SD and below during the initial months of the COVID-19 pandemic due to increased video consumption (TechCrunch, Economic Times).
Apart from bigger CDN bills for the content providers, the consumers could face problems with buffering and latency from overloaded CDNs when the consumption hits peak.
There is an alternative to solely relying on CDN/Origin-based video streaming and it’s called P2P or Peer-To-Peer video streaming. In the P2P paradigm, the video player has an option to either get the next segment of video from the CDN/Origin or from a nearby device that has previously viewed and cached the same segment of the video, thus acting as a source. P2P has the potential to reduce your CDN costs, and help you serve regions with poor edge server coverage (think Tier-2/3 cities in India).
P2P streaming is the basis of Vadoo’s business, and they are applying it to video streaming, ed-tech, and online gaming.
Let’s take an in-depth look at who they are and what they do.
About Vadoo and its Founders
Vadoo was founded by Ankur Singh and Anil Chandra Naidu Matcha, who met at Entrepreneur First, an accelerator program that brings together entrepreneurs from various fields and helps them find the right founder to pair up and build a company.
Vadoo is backed by Jio as part of the JioGenNext accelerator and are running pilots with Jio media platforms. They recently won Conquest 2020 (a startup pitch competition by BITS Pilani), winning Rs. 50 lakhs (approx $70K) in equity-free cash.
Ankur, the CEO, comes with 8yrs of experience in content & gaming in various product & business roles. He has a lot of experience in building and scaling apps rapidly (e.g., 30M users in 18 months) in the B2B space.
Anil, the CTO, has a unique skill set in building large scale networking applications and developing AI-based applications for mobile eco-systems, which is core to developing Vadoo’s proprietary technology (their secret sauce in the P2P streaming world).
How Does Vadoo’s Technology Work?
Considering that Vadoo builds P2P technology, the first thing you need to do is get Vadoo’s SDK onto your devices (clients). These SDKs communicate with Vadoo’s server-side Orchestrator that tracks a set of peers and gathers critical information about the geolocation, ISP, bandwidth conditions, network conditions, etc. of each of the peers.
The orchestrator works with a data engine to use all of this information and match different peers to enable efficient P2P streaming. And all of this is done with high-speed and performance to reduce latency for the clients.
Now, this sounds similar to other P2P solutions out there – so what makes Vadoo different?
What Makes Vadoo Different?
Considering that P2P is not a new concept, I asked Ankur Singh what differentiates Vadoo from the rest. He had the following to say –
- Vadoo has a native P2P client implementation that comes in at around 300 kb, and the size is dropping with each iteration.
- The Vadoo Orchestrator hosted on the cloud uses machine-learning at scale to identify and connect peers instead of using static rules that can get hard to manage.
- Finally, with a 4G network focus, Vadoo’s algorithms are self-learning to accurately identify and connect the right peers despite having a choppy 4G network between them.
- Vadoo uses edge-compression, which reduces both the client device’s upload and download, reducing bandwidth costs for each user. Vadoo’s AI compression engine is a self-learning algorithm and continuously monitors compression performance at the edge to self-train and improve over time.
- By using a custom super-resolution engine, Vadoo is able to deliver high-quality video at lower resolutions by upscaling them efficiently afterward. They also monitor the upscaling performance across client devices and selects the optimum neural network based on device capabilities (compute & GPU).
- Finally, Vadoo also provides a dashboard that you can use to track each client’s statistics and the total bandwidth saved. When the USP is bandwidth-savings, this is an excellent way to provide validation for your customers.
Applications for Vadoo
P2P streaming has an obvious and immediate application in video streaming, and this can help content providers reduce their bandwidth costs. P2P makes even more sense when there aren’t enough caches nearby to provide a smooth, buffer-free viewing experience. I think this has many applications for Ed-Tech, considering that online-education has become a reality and mainstream.
Vadoo has another exciting application for P2P streaming: the online, multiplayer, gaming industry. Their P2P technology can be used to add scalable voice and video engagement to their gaming platforms; and deliver gaming patches and updates in real-time.
As part of their roadmap, their solution will also enable real-time multiplayer gaming using the P2P connections between simultaneous users on browsers and native devices with significantly less latency.
Why is this interesting?
In the server-side model, individual games need to send the player’s’ position to a central server, collating and sending back information to all the players to keep them in-sync. This model has inherent problems, the biggest of which is latency. A player with lower latency has an obvious advantage in video gaming. There are workarounds for latency, such as “Lockstep State Update” in which the server does not take action until it has received inputs from all the players. However, this means that the player’s latency limits the game, which can be a spoiler for high-intensity FPS games.
Reducing latency during gameplay is something P2P streaming can help with, and that is what Vadoo is trying to achieve with its products.
Demo and Getting Your Hands on Vadoo
The simplest way to start using Vadoo is to get in touch with Ankur([email protected]) or Anil ([email protected]). They’ll set you up with an SDK that you can use for P2P streaming in your application.
If you are interested in a demo of Vadoo’s technology, then head on over to the demo on their website that allows you to start a streaming session and add peers to see how P2P fares viz-a-viz traditional ABR streaming.
Before I sign off, Vadoo’s founders wanted me to let you know that there are job openings at Vadoo. It’s a good chance to work at a fast-paced startup on some pretty cutting edge topics!
Thanks for reading, and until next time, Subscribe, Share, and take care!
Thanks to Shrinath Rao (LinkedIn, blog) for helping out with some of the research on multi-player gaming. Based out of Austin, Texas, Shrinath is an avid gamer, game developer, and is looking for his next gig.
Would you like to be featured in OTTVerse’s weekly Industry Spotlight? Send us a message and we’ll get in touch.
Krishna Rao Vijayanagar
Krishna Rao Vijayanagar, Ph.D. is the Editor-in-Chief of OTTVerse, a news portal covering technological and business news in the OTT space. With extensive experience in video compression, ABR streaming, video analytics, monetization, and more, Krishna has held multiple roles in R&D, Engineering, and Product ownership at companies such as Harmonic Inc., MediaMelon, and Airtel Digital. Krishna has published numerous articles and research papers on the latest trends in OTT and frequently speaks at industry events to share his insights and perspectives on the fundamentals and the future of OTT streaming.