Axelar AMA: From Inception to Testnet

On Tuesday, July 2nd, Axelar held its first Ask Me Anything (AMA) on YouTube, where CEO and co-founder Sergey Gorbunov answered community questions and discussed the journey from inception to testnet. Below is a full summary of the discussion as well as some questions that came from the chat that were not addressed in the AMA.

Thanks to everyone who tuned in live and submitted their questions across socials!

What was the motivation for creating Axelar?

Our team saw the next big problem for the ecosystem that we knew would become increasingly relevant as time went by. And in particular, the trend that we saw was that more and more fast and secure chains were being created for developers to build on, and while some developers wanted to go and build on top of these newer platforms, they also wanted access to the liquidity and composability of other existing ecosystems like Bitcoin and Ethereum. We realized that the need for scalable cross-chain communication is only going to grow as we’re evolving in this space. So myself and my co-founder Georgios decided to start a new venture to address this problem.

What problems in the ecosystem does Axelar address?

We are definitely entering a multi-chain world, so we’re seeing more chains being created, with the cost of launching a new chain going down. In this world, asset mobility and exchange of information across chains is becoming ever more important.

What Axelar is focusing on is building a decentralized stack to solve this challenge and connect different heterogeneous blockchains that speak different languages, have different consensus rules, and support different smart contract languages. We’re building a unified decentralized network and a protocol stack to connect and bridge different applications and platforms across all of these ecosystems.

What is Axelar’s approach to cross chain interoperability?

We want to make it very simple to plug-in any of the chains currently in existence, and we call this approach “plug and play connectivity”. At the end of the day, we don’t want to have to spend a lot of time building a connection to new chains, so we designed a system that with just a few commands through the Axelar network, we can connect to a new platform. We can also push smart contracts and deploy their node software in our decentralized network stack, and from there that platform becomes connected with all the other platforms that Axelar has interconnected previously.

So firstly, platform builders are not required to do any work, it’s a very easy integration that Axelar supports, and secondly, application builders can build on the best platform for their needs that are optimized for the features that are important to them, like fast finality or decentralization, but still leverage Axelar’s protocol for cross chain composability and asset transfer.

What makes Axelar different from other current cross-chain solutions?

Currently, there are no decentralized protocols that can scale across arbitrary blockchains. The existing solutions fall into one of two categories:

  1. There are protocols such as IBC within the Cosmos ecosystem, which is an amazing protocol and it works quite well for connecting Tendermint based chains, but has certain requirements and certain constraints. It only allows you to transfer information from one Tendermint chain based on Cosmos’ SDK to another. This protocol doesn’t scale well for existing blockchains like Bitcoin and Ethereum and some of the newer platforms like Avalanche and Algorand, so we need different solutions and we need different protocols to achieve this.
  2. We’ve seen some centralized solutions being created in the ecosystem that allow point to point bridges to go from one platform to another. These centralized solutions are pretty dangerous for the ecosystem. If you think about it, we’re taking an asset from a decentralized platform and trying to move it to another decentralized platform. Having a centralized custodian in the middle that holds the asset or who is responsible for this information transfer introduces great risk for the ecosystem.

The Axelar protocol is designed to be interoperable and decentralized so we don’t introduce additional risks for the ecosystem, while also supporting arbitrary platforms.

What are some of the use cases that the Axelar network can solve?

Some initial use cases that we’ve seen a lot of demand for are around asset transfers. If you own an asset on one of the existing blockchains like Bitcoin or Ethereum and you want to use this asset in an application that was built on one of the other blockchains, there’s no easy way to do this, and you often have to rely on a centralized solution. We’re building a network and an application on top of it that will allow users to take their assets and rapidly move them to another blockchain and use them directly in their applications, which is an initial go to market use case.

More generally, Axelar Network allows applications to compose across arbitrary chains. For instance, if you want to retrieve interest rate information from one platform, you’ll be able to execute a general request through the SSL and network and retrieve an interest rate from a different platform for your application.

What is the longer term vision for Axelar network?

Our vision is simple: We want to see a world where anyone can use decentralized applications and assets, without worrying about what network they have been built on.

In the early years of the internet, there were lots of different networks and a lot of them provided very good connectivity. However, there was a need for protocols that connected different networks together, and so players like Akamai and others built the infrastructure to more efficiently deliver content from one network to the other. Now, users don’t have to worry about varying networks and can share and communicate freely.

Axelar is building a similar value proposition for the blockchain ecosystem, except we’re focusing on routing information and assets across blockchain networks. Our mission is to build an infrastructure that allows us to bridge all these networks together, and get us to a point where we can finally start talking about interesting assets and applications and a little bit less so about different networks. Users should be able to take an asset they own and directly use it in an application of their choice, no matter what platform it was built on. And so long term, the Axelar stack has really been designed and positioned to help us get there.

How will end users interact with Axelar? Is it just builders who interact directly with Axelar or will retail users be able to interact with it as well?

There are two paths for interacting with the Axelar protocol. The original one is through the applications that either we’re building or other folks in the community are building, such as asset transfer applications. An initial application that has been built is to allow a user to take their asset and rapidly move it to another platform then participate with it directly in that platform’s applications. So in that case, users will be able to go to a web portal and use it to transfer the assets to one of the other platforms.

The second phase of integrations that we’re working on is direct integrations with application builders. You can imagine if an application builder builds on top of Ethereum, but wants to include assets of other ecosystems or compose with other applications on other ecosystems, then they will integrate their application with the Axelar protocol. By doing so, they will be able to issue requests to interact with assets of those ecosystems.

Speaking of application builders, how easy is it for a developer to integrate Axelar into their products?

For the wrapped assets that we’re bringing across the ecosystem, it’s pretty straightforward. Developers can interact with the Axelar protocol as they would interact with any other asset, like the ERC20 standard for Ethereum. And for the direct integration that I mentioned, there’s a little bit more work that has to be done, so developers will have to integrate their app directly with the Axelar protocol, but that gives them access to the whole array of the underlying assets and ecosystems that they will be able to interact with, to give them more liquidity and composability.

Speaking of ecosystem, are there any existing or upcoming DeFi projects or chains that you plan on integrating with Axelar?

We’re working with a number of pretty exciting ecosystems like Bitcoin and Ethereum as I mentioned, and we’re also working with Avalanche, Terra, Polkadot through the Moonbeam Network, and a few others that haven’t been announced yet. We have a number of projects and networks that we’re planning to integrate, so stay tuned for future announcements!

Is there anything in particular that you look for when considering industry partners like Avalanche, Terra, and Moonbeam?

What we are looking for are sets of assets and applications that we think will be interesting for the community and for the end users to interact with and use. And of course we want to work with interesting technology stacks that we think have promise and strong technical teams.

What are the latest testnet updates?

We opened up sign ups for Axelar’s incentivized testnet a month and a half ago and we’ve officially hit over 7,000 signups. Of the testers currently working through the testnet, we have 24 validator communities represented from projects like Mina, the Graph, Solana, Harmony, and many more. At the last release, the community tested how to spin up Axelar nodes and perform basic cross-chain transfers from Bitcoin to Ethereum, and at the upcoming release we’ll have three exercises available, all with various levels of difficulty.

What’s the current status of the testnet and what are the upcoming plans for it?

We’ve been running the testnet for the past few months and recently we spun it down in order to prepare for the next release, which will be very exciting for us: we’ll be inviting community members to test a number of new features. Testers will be able to join the network as validators in order to participate and validate cross-chain transfers from one chain to another.

The second really interesting feature that we’re quite excited about is the ability to plug in an arbitrary EVM chain. We’ve extended our stack so that you can execute a few commands on top of the Axelar Network, and assuming the validators have linked against that chain, then that chain will be supported by the Axelar stack and will be able to transfer assets like Bitcoin to that chain and back. This is what will allow us to start inviting our ecosystem partners to the testnet. Avalanche for example is EVM compatible so it’s pretty straightforward for us now to onboard Avalanche users to the testnet.

How long will the testnet stay open for?

The plan is to keep the test network going before the mainnet and then subsequently after the mainnet launch. Testnets are incredibly important for the community since they allow the ecosystem to start building around them and to start testing their tools. We are planning to keep it running for as long as it takes to keep on iterating and pushing more features, making it more robust, more secure etc, until the network is ready to go live.

Are there plans to open it up to the general public?

So far, we’ve been inviting folks to participate in the testnet on a rolling basis. To support more participants, we have to make sure we have the operational resources in place so that the technology can scale to support the demand, we can answer all the questions as they come up, we have enough documentation and material etc. The plan is to open it up for everybody, as it’s going to be a public network that everybody will be able to use. Currently, we’re figuring out how to surround ourselves with enough resources and support to get us there.

Keeping in mind wanting to open up for the general public, what are the technical requirements needed to participate?

First of all, we’re looking for folks that are excited about the project that we’re working on. Apart from that, folks should have a general understanding of blockchain technology, how to run nodes, and how to participate as a validator. It’s also good to have a general understanding of Docker, Linux, and Mac OS Terminal.

Why does Axelar use the Delegated-Proof-of-Stake model?

Delegated-Proof-of-Stake is a proven model that’s been working for Proof-of-Stake chains for many years. Right now, Cosmos hub, Terra, and other Proof-of-Stake blockchains are successfully using this model. In this model, you have token holders that are effectively electing a set of validators to participate and execute the consensus, and the consensus underneath it can be anything from a Byzantine agreement or another consensus protocol. This means that Axelar has a set of validators that will be responsible for executing our consensus to facilitate these cross-chain transactions, and underneath the Axelar network is an overlay of Byzantine consensus with threshold cryptography. What this model allows us to do is to optimize the network to design the correct rules for the system itself. The rules that I’m talking about are membership changes, so if a validator goes away, we have to deal with it in our network, we have to also design the correct set of incentives, and we have to design the correct set of safety and liveness rules. So the model works quite well: it’s been tested, it’s been working, and we’re designing our rules appropriately around it.

Will Axelar rely on a large network of validators?

The original goal was to start with somewhere around 100 validators and scale from there. On the technical side, there are a number of things you can do to scale the consensus and to scale the number of validators, like introducing things like BLS signatures and optimizing various consensus rules. We certainly have a team that has experience doing this, like our co-founder Georgios Vlachos, who is amazing in designing the various consensus rules and optimizing for them.

One important thing to realize for the ecosystem is that it’s not just about the number of validators on this platform, it’s also about the stake distribution. We want to make sure that decisions for the overall network are controlled by the many, and not the few. We’re designing the go to market strategy around that so that we have stake distributed across all the validators as much as possible, and we’re focusing on building a test network program and an ecosystem of partners that will also run validators.

How technical are the exercises on Axelar’s testnet?

This can vary. Some are simpler, for instance, to use the CLI commands or some of the tools that we are writing to transfer assets from one platform to another. You just have to spin up an Axelar node and run it. There are, of course, more complicated exercises, for instance if you want to join as a validator you have to do a little bit more work to set it up, and you’re going to have to know how to manage the keys for the network, and so on and so forth. We want to have exercises for everybody who’s excited about the project, and so we’re going to design them to have various levels of difficulties to account for.

Many projects have had the same ambition and failed due to technical problems, how does your project stand out and achieve its goals?

Indeed, the question of interoperability is not new in the ecosystem and people have tried to solve this for many years. First and foremost is the team that we have that we have put together as they have a deep background in cryptography, distributed systems, and consensus, and have been working on systems such as this for many years.

The other aspect is the approach that we take into designing this protocol. As I mentioned earlier, the underlying objective that we started with was to make sure we can plug in an arbitrary chain as easily as possible, allowing developers to build on any platform that they want to in order to interact with other ecosystems. We achieve this by an overlay of consensus mechanisms with threshold cryptography. The catch is that a lot of the building blocks that you need to build this were not available even a few years ago. So to give an example, if you’re building around threshold cryptography, you have to have protocols to deal with membership changes and malicious nodes in the network. If somebody misbehaves and sends a bad message across the network, you have to identify those folks and you have to potentially remove them from the consensus. You couldn’t quite do it for some of the underlying signature schemes that you needed even a few years ago. Today this technology is becoming more available and we have the right team to tackle this.

Speaking of Avalanche, how can node operators participate and help with the testnet?

As I mentioned earlier, the upcoming release of the testnet will allow us to plug in arbitrary EVM chains including Avalanche. The community will then be invited to spin up validator nodes and also spin up Avalanche nodes, link Axelar nodes against the Avalanche nodes and start moving assets such as Bitcoin back and forth. They can also test out the technology, potentially build out some of the monitoring tools around the smart contracts on the Avalanche chain and then start using these assets in different applications.

Are you planning to select members of the community to run validator nodes? If so, what are the criteria for selection?

We want to have as much participation on mainnet as possible, and want to make sure there are many validators with the stake also distributed across all of them. What would be the criteria? Well, at the end of the day, the network is open, so you spin up a validator and, assuming enough people delegate stake to you, then you’ll be selected to participate in the consensus.

Stay tuned as we update this post in the coming week with answers to the rest of the questions we received during the live!


Axelar AMA: From Inception to Testnet was originally published in Axelar on Medium, where people are continuing the conversation by highlighting and responding to this story.