Dec 14, 2015 | Alec Liu

There won’t be one blockchain to rule them all, says Marley Gray, who leads Microsoft Azure’s Blockchain as a Service (BaaS) initiative.

Instead, Marley envisions a future with countless blockchains, constantly evolving and iterating, and all interconnected through a standardized protocol.

Marley is the director of technology strategy for financial services in the US at Microsoft. He also runs the company’s innovation labs in New York City. Microsoft made a splash recently when they announced their first BaaS offering in Ethereumon their Azure cloud platform. Then last week, Microsoft unveiled a partnership with Ripple.

The following interview has been edited for clarity and length.

It’s obviously super exciting when a company as influential as Microsoft decides to dip their toes in a new technology. Is the company’s interest in blockchain technology a recent development?

We’ve been looking at blockchain in our innovation labs for quite some time, such as fooling around with Colored Coins, stuff like that. But what was most compelling to us was smart contracts since it’s not just about payments. So we decided to check out Ethereum. Working on the early build with that platform—it was painful to say the least.

This naturally presented itself as an opportunity since our customers looking into these technologies are going through that same painful process. There had to be a way to streamline this whole process, create a fluid environment where you can quickly spin out the required infrastructure with all its dependencies and do it in the cloud—and to be able to do that with a single click. That’s essentially how blockchain as a service was born. We went from having to go through this 2 day ordeal of setting up a lab for a blockchain experiment to being able to run create a blockchain in the cloud with a single click. You can have your first smart contract up and running within 20 minutes.

Ethereum is first out of the gate but we have a whole slew of other blockchains in the mix. We actually went to Ripple first early on in the process but we weren’t sure at the time how payments fit into our model. But as we started learning more about how all these different pieces fit together in the ecosystem, Ripple starts to make a lot of sense especially when you throw the Interledger Protocol into the mix. At the end of the day, the thing underlying all of these applications and smart contracts is payments.

But ultimately, we’re blockchain agnostic. Our mission is straightforward—we’re looking at the marriage of hyperscale distributed cloud platforms that are open source. That means we’re open to any OS and any framework really. Does it scale? Is it fast? Is it easy to learn? Does it lend itself to this kind of environment? We can figure that stuff out a lot cheaper in the cloud than people having their own datacenters. Right now, we’re working through the different options and we’re still doing a lot of experiments with architecture.

These are still early days but who do you foresee as your primary customers for BaaS? Is it mainly developers?

It’s funny, the initial touchpoint in the financial services vertical is generally the developer, where you have developers aligning with business people especially in capital markets, which is becoming increasingly technologically oriented. You see this a lot with trading desks where they have their own developers we come along to all the meetings. The desk will be business led but a lot of underlying innovation will be managed by the developer. So for us, the DevOps community is the first play.

Then there’s this whole tooling system that has to be built up. The objective is for people to be able to model contracts using a visual system, like Microsoft Office, for instance, where someone can create a smart contract by dragging and dropping different functionalities and setting different rules through a user-friendly interface. That way, the business guys can build something functional around an idea they have where it’s not just some throwaway modeling. They can pass it on to the developer, who then puts meat around the scaffolding. The final vision is a complete end-to-end service. We’re looking at the whole stack for building next generation applications.

Was this a company directive or a personal mission for you?

It started out as a more personal thing. I’m the technologist on the team. The rest of the people around me are investment bankers. I came from a place, earlier in my career, where I was running trading floor applications. Given that experience, I was intrigued by the blockchain’s potential. Once I started talking about it internally with my colleagues it started to click and catch, especially at the financial centers like New York and London. It sort of grew organically from there. Right now, it feels like we’re experiencing exponential growth. We’re very excited about this project.

What’s your take on the state of blockchain right now?

Right now, there’s a trend toward forming consortiums around various settlement use cases, which I think will play a big role, especially on Wall Street, such as with R3 and the commodity consortium we’re participating in. These consortiums will help drive collaboration, hone in on the needs of their stakeholders, and figure out what technology they need to adopt or develop.

From Microsoft’s vantage point, we’re focused on infrastructure. We’re not looking to write our own blockchain stack. We want to support the innovation that’s already out there and help build a robust ecosystem—like our investments in Docker and other container technologies. The end-game is to enable plug and play at the development level so people can start iterating these technologies much more rapidly.

Because this is just the beginning. Right now we have Bitcoin, which is this public network. There will be people out there who will tell you that you need to be on the Bitcoin blockchain. We disagree. For a couple of reasons, we don’t really see it going anywhere. But Bitcoin served an important purpose. It proved that the technology can work. But it’s just one chain. We’re barely scratching the surface.

In financial services, I think we’re going to see a mixture of different blockchain technologies. There will be many, many chains—chains for swaps, chains for bonds. They’ll each have different characteristics. We’ll see chain patterns evolve. You could plug and play various consensus algorithms based on the products that exist on that chain.

Given this multi-chain vision, we see smart contracts as a key piece. Then you need ILP, which is the glue in the system that holds everything together, so that different chains can interoperate. That means a derivatives chain can interact with the securities chain it’s derived from. In all of these interactions, you’ll have to move money around. So you will need Ripple for that.

We see Ripple and ILP becoming part of the fabric of this multi-chain reality. Then you have Azure, which is the horizontal computing layer. We’ll provide scale management. We’ll make sure there’s regulatory compliance for data residency. We’ve got all that stuff covered.

The fact that Microsoft is running a Ripple validator has our offices buzzing, naturally. How’s the experience been so far on your end?

It’s been great. It’s actually really easy to do. We initially made this big virtual machine but realized it wasn’t necessary so we scaled it back down. Overall, the performance has been very good. We’re currently going through the process of getting verified. Eventually, we’ll probably run more than one validator.

• See Ripple’s validator registry

We want to be the premier validating node on the Ripple network. We believe in Ripple so we want to put our full weight behind it so that the ecosystem develops in the right direction. We’re really excited about the partnership. I have a bunch of meetings scheduled with the rest of the team to dive deeper into the different kinds of stuff we can do with Ripple and ILP.