Angela Ramirez (YC 12), Software Engineer at Livepeer, Ethereum-based Video Encoding Platform
Angela Ramirez is a Software Engineer at Livepeer, a decentralized platform for video encoding built on the Ethereum network. Ms. Ramirez works on the Golang implementation of the Livepeer protocol, which allows anyone to create a local Livepeer Network, broadcast a video stream into the network, request a video stream to be transcoded, or consume a live video stream from the network. After graduating from the Flatiron School’s Full-Stack Web Development course, Ms. Ramirez briefly worked as a Full-Stack Software Engineer at Nimblestack and later moved to BuzzFeed, where she worked as a Full-Stack Software Engineer on the Video Production team for two years before joining Livepeer. Ms. Ramirez received a BA from Yale University in Political Science and International Studies.
The Politic: Tell me about your background and how you got involved in blockchain!
Angela Ramirez: I came to Yale with a very strong interest in becoming some kind of engineer at NASA. But when I got to Yale, I was fascinated by the political science, history, and philosophy classes. These were subjects that I’d never been exposed to, and so I ended up majoring in Political Science and International Studies. I wasn’t in tech at all. I wrote for The Yale Globalist, and I traveled the world. I did a bunch of projects in India, Jordan, Tanzania, and so on.
After graduating from college, I knew I wanted to do something good for the community and go back into a quantitative-intensive job. I did TFA for two years after graduating and taught history to sixth graders. Then, I went on to a boutique, non-profit investment bank, where I did the more quantitative work of building financial models and preparing investment recommendations.
Deep down, I always knew that I wanted to be involved in tech, and that I wanted to go into engineering. I knew that coding boot camps were a thing, so when I was at that small bank for two years, I spent the last year doing a lot of tech work. I was working on the website, working on the Salesforce systems, and really doing anything that would have otherwise been outsourced to tech companies or consultants. I saved my company a lot of money but also gained technical experience I knew I would need in the future.
Then, I did Flatiron’s coding school for three months, and I’ve never worked so hard in my life. After that, I went to work for Buzzfeed on their video infrastructure team. That had to do with any of the Buzzfeed videos that you see on their website, Facebook, YouTube, Instagram, or any other platforms where they output video. We had a transcoding pipeline for videos, so I rebuilt that pipeline to be in line with modern requirements, and after that, I was very interested in the backend software engineering side of things.
The current company that I work for, Livepeer, reached out and said that they needed someone to do that work. When I joined, there were 9 people. Now we have about 15 people. We’re an open source video processing services start up built on the Ethereum blockchain. We leverage decentralized technology that makes possible use cases and business models that were never available before. It’s a very different use case for the blockchain; I haven’t seen anyone else working on what we’re doing. We use the blockchain to create a world-wide marketplace for video transcoding.
Individuals who hold Livepeer Tokens can stake them toward their favorite orchestrators (these are nodes that coordinate transcoding work), and in doing so, encourage active participation in the network and incentivize teams around the world capable of running transcoders to actually do so. This makes the network increasingly stable and performant.
Our token is listed on exchanges, but we’re not pushing towards that. So far, more than 100 events have been streamed using our software. We use the Ethereum blockchain for payment exchanges, transaction security, and worker discovery. We recently raised $8 million in Series A funding, and we’re working towards our goal of Livepeer 2.0.
What’s video transcoding, and how does it work?
Our network runs using different nodes, and we have three nodes that do the work of video transcoding. If you’ve seen the show Silicon Valley, you kind of understand what we do. Video first goes through our system through a broadcasting node, then an orchestrator node that’s responsible for distributing jobs to many different transcoders. The video is then transcoded by an available, capable transcoder node, and is sent back to the Broadcaster for live streaming
Transcoding is taking a video and making different copies of that video at different bitrates, heights, widths, and so on. So, you might have a video that’s coming out of someone’s camera at a 4K-resolution, but that would be incredibly expensive to send over the internet. It’s not feasible for all high definition videos that are out there to be sent intact through the internet – there wouldn’t be enough broadband available for all videos in the world to be transmitted given current technological limitations. Plus, it would be very expensive to do so.
Instead, what we do is help video broadcasters find a capable transcoder who can compress and decompress that video. There are different techniques, or “codecs” you can use to do so, and a host of other transcoding options to take advantage of. Livepeer makes it a seamless experience, saving more than 10 times the cost you’d pay to a centralized provider like AWS.
At BuzzFeed, we would send our high-resolution video to AWS, and AWS would take care of doing the transcoding and sending back different renditions, i.e., different versions of that video at lower resolutions, the most common being 360p, 480p, 720p, and 1080p. So, we would send that video from BuzzFeed to AWS, and they would send that video back to us at BuzzFeed.
The way that Livepeer is different is that we’re not a centralized provider. What we really do is facilitate the interaction between anyone, anywhere in the world, assuming they have extra computing power to perform some of this computationally-intensive transcoding work. We make sure they’re connected and able to communicate by using smart contracts on the blockchain. So, those two parties can communicate in a trustless and secure way.
What’s a concrete example of one of your video partnerships?
Take the example of BuzzFeed. Instead of using AWS, they could send their video to one of our broadcaster nodes, which would then go to an orchestrator node, each of which are run by individuals anywhere in the world. Broadcasters send micropayments with segments of their video stream that compensate orchestrators for transcoding the segments. In case of any wrongdoing, we have a system of “slashing” a node’s collateral or deposit on the blockchain. This is an incentive mechanism for people to do the work correctly and act in good faith.
On the transcoding provider side, we‘re giving individuals who have either been miners or have extra computing capacity lying around, an opportunity to earn revenue from their hardware. As blockchains shift from proof of work to proof of stake, there’s a lot of freed up computing capacity that is now unused. Miners, or any other individuals with server space, now have an opportunity to make money transcoding, something that no other company is offering.
Our network is not immune to the laws of supply and demand. As we grow and more individuals around the world become a part of it, the price of transcoding, we predict, will settle at potentially even lower levels than seen today. So, if there’s a lot of demand for transcoding, and not as much supply, prices will likely increase. In the opposite scenario, prices could reach new lows. It’s an economic system, and we theorize the cost will be more than 10 times less than what it costs right now to do video transcoding on AWS.
That’s important today because 80 percent of all internet bandwidth is consumed by video, and this is really expensive and hard to execute. Think of YouTube or Facebook. They have their own transcoding systems. They could cut their cost of video transcoding by more than 10 times for certain kinds of content. Centralized providers like AWS profit from this industry by marking up their transcoding resources and their computing resources by a lot. Hopefully, our tech will completely change that and force those providers into a competitive market where they’re up against any individual, anywhere in the world.
We believe that people in countries where electricity is really cheap, for example in Iceland, will provide transcoding resources at barely a little bit above the cost of electricity. That’s going to be a lot less than what you would pay to AWS or any other centralized providers.
We’re currently trying to include the millions of GPUs that have been bought by crypto miners over the past few years in our marketplace. There’s a lot of idle computing power out there. It’s like if you think about Amazon, but Amazon for transcoding services. We’re essentially creating a network on top of that excess capacity and allowing people with video cards and knowledge to run software on their computing systems to participate.
At the moment, we’re targeting a lot of the user-generated content and other long-tail content. YouTube has their own set up. Facebook has their own set up. They could each call us and say, “Hey, we want to use Livepeer” for some of that content. However, given that 80 percent of traffic on the internet is video, there’s a huge opportunity outside of that for anyone to come into the market and really radically reduce the price of transcoding.
Can you describe your day-to-day environment?
On a day-to-day basis, because we’re a small company with a very tech-heavy, software-heavy product, most people who work for Livepeer are engineers. We only have three people who are not engineers, and those people are pretty technical themselves. Most of what we do now, given where we are in our development as a company and given that our product is still in development, is calling different video-related companies, making them aware that Livepeer is an option. We don’t have customers yet, we have design partners, but we have the funding to develop for the next couple of years. So, we try to hear what these companies think of our product, and we try to hear about different requirements in their tech set up. Then, we use that information to create product requirements, and do some of the coding to make them a reality.
At a bigger company, that might be handled by three different people at the same time: those who are in charge of customer development, product management, and engineering. But, the fact that we’re so small requires us to wear a lot of different hats. There might be a day, like tomorrow, where I don’t have any meetings, so I’ll spend the day glued to my screen working on specific features of the code. That being said, you’re the eighth call of my day, so it really just depends.
I think what people should know is that it’s just so exciting working for a small company, in a startup that’s in such a new industry, that’s working on a problem that no one else is trying to solve. Every day is different. Every day, you’re talking to a different person in a different part of the world. That really emphasizes the role of having a community, and an international one at that, in the blockchain space. A lot of these calls are with people in Europe. Today, I had a call with someone in London, a call with someone in Berlin, and call with someone in Kiev. We have teammates all over the world as well. We’re not just a company using decentralized technology but also a decentralized company. Location really doesn’t matter.
Any final words for the readers?
I think it’s important for people to both look at the crypto side of things but also the fact that we’re a startup. I would highly recommend working at a place like Livepeer because you get to touch a lot of different parts of the company, and even though you might have a much more stable and predictable schedule at a larger company, things change very quickly for us. I find it exciting. That said, It’s not for everyone. Some people prefer a more regimented environment where they know what they’re going to do every day, for the next month or more, and that might actually be a good fit for somebody who’s just graduating.
But, I think it’s important for people to know that there’s a whole world out there and a lot of different options in the blockchain industry. You can almost picture the connections between people all over the world in your mind. Because of that, my team and I go to conferences all the time. In two weeks, for example, I’m going to Amsterdam for a video conference. Last fall, I went to San Francisco and Prague for a couple of others. I have teammates who will likely fly to Japan later this year for something more focused on Ethereum, and have had teammates going to Buenos Aires, all over Europe, Mexico, Asia, etc. for conferences in the past. It’s a small industry compared to other industries, but that just makes it all the more exciting. I love it. I love my job. It’s a great opportunity to come up with fresh ideas and make a difference.