Kevin Hurley

So it was really hard for customers to understand who do we open channels to? What are channels even mean? How do you get inbound liquidity? How do you make this work well? So we found that the first version was still too complex.

So it was really hard for customers to understand who do we open channels to? What are channels even mean? How do you get inbound liquidity? How do you make this work well? So we found that the first version was still too complex.

We quickly went back to the drawing board and tried to iterate on that and get it to something that was even simpler and that kind of abstracted the notion of channels entirely.

We quickly went back to the drawing board and tried to iterate on that and get it to something that was even simpler and that kind of abstracted the notion of channels entirely.

We quickly went back to the drawing board and tried to iterate on that and get it to something that was even simpler and that kind of abstracted the notion of channels entirely.

Like I mentioned, we started out by running a full L&D node for each customer. which simplified our stack because that's how L&D is packaged today. But obviously that wasn't something that would be scalable or efficient long-term. If you have to spin up an AWS pod for every single customer, if you have hundreds or millions of customers, that's obviously not going to scale well.

Like I mentioned, we started out by running a full L&D node for each customer. which simplified our stack because that's how L&D is packaged today. But obviously that wasn't something that would be scalable or efficient long-term. If you have to spin up an AWS pod for every single customer, if you have hundreds or millions of customers, that's obviously not going to scale well.

Like I mentioned, we started out by running a full L&D node for each customer. which simplified our stack because that's how L&D is packaged today. But obviously that wasn't something that would be scalable or efficient long-term. If you have to spin up an AWS pod for every single customer, if you have hundreds or millions of customers, that's obviously not going to scale well.

So we started off taking that as we want to get something out on the market. We want to be able to prove that we can actually simplify how Lightning works and make it so that customers can easily integrate. And from there, we'll be able to figure out how we scale better and actually make the right trade-offs later on.

So we started off taking that as we want to get something out on the market. We want to be able to prove that we can actually simplify how Lightning works and make it so that customers can easily integrate. And from there, we'll be able to figure out how we scale better and actually make the right trade-offs later on.

So we started off taking that as we want to get something out on the market. We want to be able to prove that we can actually simplify how Lightning works and make it so that customers can easily integrate. And from there, we'll be able to figure out how we scale better and actually make the right trade-offs later on.

But we didn't want to overcomplicate things at the start until we actually felt like we had product market fit and that we had something that was going to be valuable. I think another area that we had a trade-off was that we really just overcapitalized our nodes initially.

But we didn't want to overcomplicate things at the start until we actually felt like we had product market fit and that we had something that was going to be valuable. I think another area that we had a trade-off was that we really just overcapitalized our nodes initially.

But we didn't want to overcomplicate things at the start until we actually felt like we had product market fit and that we had something that was going to be valuable. I think another area that we had a trade-off was that we really just overcapitalized our nodes initially.

So until we actually understood the network better and could see the usage patterns of how our customers were using the network, the easy solution was just to throw more capital at the problem, make it so that we could make sure that we can route whatever transactions our customers might have and ensure they have a good experience no matter what their usage patterns might be.

So until we actually understood the network better and could see the usage patterns of how our customers were using the network, the easy solution was just to throw more capital at the problem, make it so that we could make sure that we can route whatever transactions our customers might have and ensure they have a good experience no matter what their usage patterns might be.

So until we actually understood the network better and could see the usage patterns of how our customers were using the network, the easy solution was just to throw more capital at the problem, make it so that we could make sure that we can route whatever transactions our customers might have and ensure they have a good experience no matter what their usage patterns might be.

Now that we've been running for a while, we can much more efficiently manage that. Once we've seen the usage patterns and once we've been able to understand better how customer funds might flow, we can make sure that we're actually connected throughout the graph really well without having to overcapitalize.

Now that we've been running for a while, we can much more efficiently manage that. Once we've seen the usage patterns and once we've been able to understand better how customer funds might flow, we can make sure that we're actually connected throughout the graph really well without having to overcapitalize.

Now that we've been running for a while, we can much more efficiently manage that. Once we've seen the usage patterns and once we've been able to understand better how customer funds might flow, we can make sure that we're actually connected throughout the graph really well without having to overcapitalize.