Sep 24, 2024

LiveKit Reveals The Internet Wasn’t Built for Us

An interview with Russ d'Sa, the founder of LiveKit

Founder Focused

Raising $85M? More than That

Russ d'Sa Speaking
Russ d'Sa Speaking
Russ d'Sa Speaking
“When you go to the address bar, you enter HTTPs. That's Hyper Text Transfer Protocols. It was built for transferring text."
Russ d'Sa
Founder of LiveKit
This isn't just another startup story. It's about a serial entrepreneur who has consistently chosen to walk away from the comfortable path to pursue bolder visions. It was time for someone to stand up against those limitations.
"As AI becomes more human-like, we will need new infrastructure to support natural interactions through sight and sound."
Russ d'Sa
Founder of LiveKit
Today, LiveKit powers 25% of America's 911 dispatch centers, enabling life-saving video assistance during emergencies. The company raised $85M funding in total with participation from Altimeter Capital and prominent AI industry leaders like Google Chief Scientist Jeff Dean.
But more importantly, Russ d'Sa is positioning LiveKit to become the foundational infrastructure for how humans will interact with AI in the future

Current Internet Is Made For Text

This limitation becomes particularly apparent when trying to build applications that require real-time interaction. d'Sa encountered this problem firsthand during the pandemic while working on a side project that required real-time audio capabilities.
“Everybody was bringing different elements of real life onto the internet. They needed a way to use cameras and microphones to do that."
Russ d'Sa
Founder of LiveKit
The traditional internet protocol requires constant back-and-forth confirmation of data receipt - a "handshake" that creates noticeable delays. While this works fine for text, it becomes problematic for high-bandwidth data like video and audio where smooth, instantaneous transmission is essential.
d'Sa and his co-founder, Zhao are testing LiveKit, ©LiveKit
d'Sa and his co-founder, Zhao are testing LiveKit, ©LiveKit
d'Sa and his co-founder, Zhao are testing LiveKit, ©LiveKit
"I pinged my co-founder David and said 'Hey, what if we build the general purpose infrastructure that makes really easy to build any kind of real-time audio or video using application?'"
Russ d'Sa
Founder of LiveKit
Existing solutions lacked reliability. While companies like Zoom have built internal stacks to handle real-time media, there was a void in a field of open-source alternatives.

Why Does Fixing Infrastructure Hard

Building LiveKit presented unique technical challenges. The team had to create a software-defined routing system that could handle global scale while maintaining latency under 200 milliseconds.
“The world of the internet is like the road system, Sometimes you might get on a congested road. Sometimes it’s broken. That's what the internet looks like."
Russ d'Sa
Founder of LiveKit
d'Sa depicted that "the hardest part" is you don't know what the internet looks like until you actually try and traverse some of those roads.
Rather than trying to plan for every contingency, he embraced a philosophy of learning through real-world usage. This meant the team needed early adopters willing to weather growing pains.
"We faced a lot of outages. I had a lot of very angry early customers. But also developed a strong relationship with those customers as we were building. I think what helped is that they were founders themselves. They understood the process of building something new, get it off the ground."
Russ d'Sa
Founder of LiveKit
d'Sa Talking to EO Producers
d'Sa Talking to EO Producers
d'Sa Talking to EO Producers
This approach led rapid improvements in the platform's reliability.
“Having your system fail and crash is actually a rite of passage. It's proof that you've built something worth improving. Any code worth writing is worth writing twice or thrice." d'Sa argues."
Russ d'Sa
Founder of LiveKit

Post-Keyboard Era

While LiveKit started as a solution for video conferencing, the team soon discovered a far more ambitious opportunity - becoming the fundamental infrastructure layer for artificial intelligence.
As AI capabilities expand, d'Sa sees LiveKit playing an increasingly central role in how these systems interact with the world.
"If OpenAI, Anthropic, Gemini and the foundational model companies collectively are going to build a brain - they're going to build a human-like computer that can respond to you like a human does.”
Russ d'Sa
Founder of LiveKit
d'Sa draws a compelling parallel between human biology and the future of human-AI interaction, while asking the critical question.
“How will you interact with that human-like computer? You will probably interact the same way that you interact with another human."
Russ d'Sa
Founder of LiveKit
The native input and output for a human are eyes, ears, and mouth.
The eyes of a computer are a camera.
The ears of a computer are a microphone.
The mouth of a computer is a speaker.
The market potential is staggering.
“Every keyboard and mouse is our market. Because every keyboard and mouse will be replaced by a camera and a microphone."
Russ d'Sa
Founder of LiveKit
This vision has already begun to materialize.
When OpenAI launched voice interactions for ChatGPT, they used LiveKit to power the real-time audio transmission.
"If OpenAI is building the brain, LiveKit is building the nervous system to carry signals to and from that brain."
Russ d'Sa
Founder of LiveKit

Left Twitter, Because Time Matters

Beyond LiveKit, there was d’Sa’s journey lying there, alternating between founding startups and joining established companies, including stints at Twitter and 23andMe.
Before LiveKit, he founded four startups, including Evie Labs, which was acquired by Medium for $30M.
©Twitter
©Twitter
©Twitter
“When I joined, there were 80 people. Two years later, we were 1,500."
Russ d'Sa
Founder of LiveKit
Shaking his heads, d'Sa recalled. 
"Things that used to take days started to take months."
Russ d'Sa
Founder of LiveKit
The realization hit him; despite the generous compensation package waiting, he needed to leave.
"I could have stayed and made a lot of money. But the only non-renewable resource in life is time. You can make more of everything else, but not time. I wasn't maximizing the most valuable being at Twitter."

"That’s the only advantage a startup has, speed. Bigger companies will always have more money, more people, more resources. But they can't move as fast as a small, focused team."
Russ d'Sa
Founder of LiveKit

Be a Late-Starter (especially in the AI era)

"AI and robotics are going to be the largest spaces in the history of humanity. They're both guaranteed to be multi-trillion-dollar spaces. But they're also so broad. How do you build a product that can grow as AI and robotics are adopted and proliferate?"
Russ d'Sa
Founder of LiveKit
His advice? Use human capabilities as a framework for understanding AI's trajectory.
“Think about what humans can do, and AI models. Ultimately the destiny of AI is to build a synthetic human being."
Russ d'Sa
Founder of LiveKit
If you could create a synthetic human, what would remain beyond its capabilities and what limits do real humans still face today? This transition to more natural interfaces - speaking, gesturing, showing - represents a massive opportunity.
Quadrant of AI Development
Quadrant of AI Development
Quadrant of AI Development
This framework helps identify both current limitations of AI and opportunities for complementary tools and infrastructure.
d'Sa also suggests that in rapidly evolving fields like AI, there can be advantages to joining established companies before starting your own venture.
"There is some benefit, counterintuitively, to waiting a little bit. First, Join a company that is at the forefront of AI. You can learn about the foundational elements, how fast things are moving, and also meet co-founders. Then, leave those companies, go and start the next wave."
Russ d'Sa
Founder of LiveKit
Russ d'Sa Speaking
Russ d'Sa Speaking
Russ d'Sa Speaking
“Do not start a company, for the sake of starting a company."
Russ d'Sa
Founder of LiveKit
He emphasized the importance of conviction. "In my early days, I easily forgot to focus on what my idea was.”
Feedback is important, but what really matters is why you’ve decided to find a company from the first place.
“Yes, all of these founders are very smart, but they're not thinking about the problem that you're solving 24/7, like the way you are."
Russ d'Sa
Founder of LiveKit
Today, Livekit is enabling educational platforms in developing countries, including NGOs in India to help students access global educational resources.
For d'Sa, making a lasting impact on technological future is more important than just a business opportunity.
“These are things that don't actually even pay us a dime. They all use LiveKit open source, but still my favorite use cases. They truly affect the world and move things forward, change humanity."
Russ d'Sa
Founder of LiveKit
As artificial intelligence continues to reshape our world, d'Sa and LiveKit are working to ensure that the infrastructure connecting humans and AI evolves to match our natural ways of interaction.
Be the synapses that bridge these two worlds, one video stream, one voice command at a time. It's a bold vision that challenges the internet's status quo. Exactly the kind of rebellion that has driven technological progress.

Friction Point

LiveKit isn’t just rebuilding internet infrastructure. It’s suggesting a new frame of how humans and machines interact. Even though today’s internet can technically handle real-time audio and video, the gap between what’s “possible” and what feels truly seamless keeps widening.
But innovation raises its own questions...
Audio/Video Infrastructure = Real Priority?
The Next Privacy Fallacy?

What To Carry

1. Don’t Build for Yesterday’s Internet.

Real-time interaction breaks down when built on protocols designed for text, not voice or video.
“The internet wasn’t really built for transmitting audio and video high bandwidth data… it was built for transferring text.”
You can’t create instant, seamless communication with tools made for email and hyperlinks. Founders solving for the next interface wave - voice, video, and AI - need infrastructure that reflects today’s reality, not yesterday’s browser.

2. Speed Isn’t Optional. It’s the Only Edge.

D’Sa left high-paying roles because speed mattered more than salary or scale.
“The only non-renewable resource in life is time… I wasn’t maximizing the value of my time being at Twitter.”
Resources are replicable. Time isn’t. In fast-moving markets like AI or infrastructure, slow teams don’t survive. Build fast, learn fast, or watch someone else do it.

3. Don’t Just Chase AI, Enable It.

LiveKit isn’t an AI model. It’s the nervous system that lets AI interact with the world like humans do.
“Every keyboard and mouse is our market, because every keyboard and mouse is going to get replaced by a camera and a microphone.”
Everyone’s racing to build “smart.” But the real revolution is in how intelligence connects. Infrastructure for sight, sound, and speech is what will power the post-keyboard era.

Join the 1.5M+ founders inbox
to get the latest updates.

Explore more