I recently spoke with Ayush Ranjan, Co-Founder & CEO at
Can you please share the origin story for Huddle01 and give us some insight into what you do?
Sure, Huddle01 started at ETH Global HackFS, it was a 30-day hackathon, and this was July 2020 - the peak time of video conferencing and Zoom stocks were through the roof with the enormous usage it was getting.
Though the stock prices of real-time communication apps & infrastructures were soaring, so were instances of invasion of privacy and general laggy performance. These were a couple of the triggering reasons which led us to dabble into the world of solving real-time communications by building a people-powered communication network.
The name Huddle01 came into place for a simple reason, people will be huddling in a digital/virtual environment, and since digital is made up of binary i.e., 0s and 1s, that’s why Huddle01!
Building a video meeting dApp which solves onboarding, matches the scalability of Zoom and Google Meet, and has similar product feature parity to these centralized web2 apps while making it possible to have new emergent behaviors on the dApp because of use of web3 primitives was extremely challenging on product, technology, and research levels.
In the last three years, Huddle01 has built a suite of audio and video products (video meetings and audio spaces products), plug-and-play SDKs and has built its dRTC network, which will be live next quarter.
In the Huddle01 dApp and Huddle01 Audio Spaces products, you can do all the things that are possible on Zoom/Google Meet and Twitter Spaces/Clubhouse, respectively, and much more which isn’t possible on these products. For instance, it is possible to do multi-live streaming via Huddle01 on YouTube, Twitch, and Twitter, do token-gated meetings, use decentralized identities, and store your recordings over IPFS and Filecoin, amongst other things.
It recently crossed 1 million minutes of meetings on its infrastructure, with 150+ dApps using its infrastructure for real-time audio and video purposes.
Zoom's recent changes to its terms of service, allowing the firm to use its customers’ data for training its AI algorithms, have caused some furor among privacy enthusiasts. How can decentralized alternatives to platforms like Zoom avoid these issues?
Well, the current version of Huddle01 uses Base WebRTC as its protocol to understand the nuances and difficulties associated with it so as to build our Node-Node Infra.
WebRTC uses DTLS/SRTP as its protocol for security, where SRTP is responsible for encrypting various types of data, like audio and video, that are transmitted through a WebRTC session. If unauthorized individuals intercept this data during transmission, they won’t be able to understand it without the proper decryption key. DTLS-SRTP is the method used to securely exchange the decryption key among devices engaged in WebRTC communication.
This ensures that participants at both ends of a call can decrypt and access the shared audio, video, and other data. DTLS-SRTP includes its own security measures that enhance protection against interception of both the decryption keys and the data itself. Despite WebRTC being a peer-to-peer service, it commonly involves connecting through a web server that manages connections between participating devices. This server’s role is mainly about facilitating communication by sending signals that enable devices within the WebRTC session to locate each other on the internet and establish a connection.
At Huddle01, we are actively exploring the implementation of insertable streams. This entails enhancing client capabilities to transmit highly secure and encrypted streams through our global RTC transports. Currently, this feature is available upon request, but our plan is to make it the default option as soon as we can ensure the stability of the decryption mechanism and bolster security measures such as key rotation.
Technically speaking, how can the development of a Decentralized Real-Time Communication engine ensure secure end-to-end user encryption as well as assist in the creation of a bottom-up economy?
By building on peer-to-peer protocols like WebRTC and Libp2p, each user's communications are encrypted between their device and the devices of other participants. Media streams (audio, video, etc.) are encrypted at the sender's device before being sent to peers. This uses WebRTC standards like SRTP to encrypt streams with AES encryption and HMAC keys.
Encryption keys are exchanged between users directly using WebRTC's DTLS protocol. This establishes a secure peer-to-peer channel for key exchange separate from the media stream. Keys are generated locally on each user's device; they are not managed by any centralized server. This prevents keys from being accessed by third parties.
The nodes only relay encrypted streams; they do not have access to decryption keys. This provides privacy from the infrastructure layer.
The network uses a consensus of nodes to relay and verify communications between peers. These nodes earn crypto token rewards for providing reliable service to the network. The more nodes that join, the more decentralized and resilient the network becomes.
There are low barriers to becoming a node - anyone with a computer and internet connection can spin up a node. This is unlike traditional centralized infrastructure that is controlled by large tech companies. Validators are incentivized to provide quality service so they can continue earning rewards. More validators mean more competition, which should improve service quality.
End users also pay lower costs since they aren't paying steep fees to centralized middlemen. The microtransactions for using the network can be very affordable. Developers can build apps/services on top of the open dRTC protocols. This further expands economic opportunities on both the infrastructure and application layers.
In light of the growing public scrutiny over AI and data privacy, how does a decentralized approach provide a solution?
Huddle01's decentralized architecture is uniquely positioned to address the growing public concerns around AI ethics and data privacy. By avoiding centralized servers, Huddle01 prevents any single entity from accessing users' communication data or surveilling activity across the network. This aligns with the current demands for better data protection in a few ways. First, end-to-end encryption ensures user communications remain private and accessible only by the intended recipients, meaning no third parties can access call contents.
Next, nodes operate on an encrypted data plane, merely relaying information without visibility into the contents. Additionally, dRTC protocol allows transparency into how the network functions by open source standards, meaning no "black box" algorithms. Finally, users retain control over their information and interactions within the network rather than relying on centralized providers' policies.
Considering the ongoing conversation surrounding today's data privacy laws, what are your thoughts on the legal landscape surrounding this space?
In the current architectural landscape of major Real-Time Communication (RTC) providers, their systems are considered highly secure due to their centralized control, which gives them exclusive command over the infrastructure. However, the peer-to-peer RTC protocol, WebRTC, also boasts strong security and privacy features through the use of DTLS/SRTP. Despite its robustness, WebRTC faces limitations in scalability.
As a solution, companies are transitioning towards a Client-Server Model, leveraging its potent control capabilities and inbuilt security protocols like DTLS/SRTP for securing communication between clients and their servers. However, this approach encounters challenges when it comes to maintaining security on the server side as TLS has to be broken to read the payload to create AI models.
Huddle01 is creating a Node-Node Architecture akin to what IPFS/Filecoin employs. In this innovative paradigm, centralized control is relinquished in favor of a protocol-driven network of interconnected nodes. These nodes collectively oversee processes ranging from node selection and operation to incentivization, mimicking the functions of a Client-Server Architecture without a central point of authority.
This transition also facilitates the incorporation of security measures such as End-to-End Encryption (E2EE) and Transport Layer Security (TLS) for RTC transports, aspects feasible in both the Client-Server and Node-Node Models.
In essence, Huddle01's approach preserves the same protective mechanisms found in the Client-Server Architecture while decentralizing control through a Node-Node Architecture, eliminating the concept of a central locus of control. Making the Peer to Peer Network, which is both secure and private, also scale using the advancements in Peer to Peer Systems.
The concept of decentralized real-time communication is relatively new, albeit promising. How do you see it impacting the broader tech industry?
Decentralized Physical Infrastructure networks (DePIN) are the next step in the evolution of crypto-based networks. The DePIN sector hosts protocols aiming to disrupt traditional, billion-dollar infrastructure industries, which can be broken down into four main categories: Server (Cloud) Based Networks, Wireless Networks, Sensor Networks, and Energy based networks. Huddle01’s dRTC (decentralized real-time communication) network, which will democratize WebRTC protocol comes under the cloud-based networks, which will be people-powered rather than relying on centralized infrastructures.
Similar to how Helium Protocol has democratized LoRaWan & 5G protocols in the category of wireless networks, Huddle01 will democratize WebRTC protocol in the category of server-based networks. As more and more protocols disrupt the centralized billion-dollar industries, we’d be seeing more people-powered networks in everyday life, and Huddle01’s dRTC will play an important role as humans rely on communication to collaborate and innovate.