Are we there yet? Understanding the complexities of blockchain production
It's getting to be weekly reading — the latest news release announcing that Bank A has successfully carried out a blockchain proof of concept with Bank B, typically involving a cross-border transaction of some type.
Over the past few years, these news flashes have generated waves of excitement about the brilliant future of distributed ledger technologies within the financial sector.
However they have not generated the sea change in financial ecosystems hinted at — if not outright promised — in almost every announcement. So, what's the holdup? Why is it taking so long to get from proof of concept to commercial product?
A half-dozen blockchain project veterans took up this question during a panel discussion titled, "Are we in production yet?", at the Blockchain and Digital Currencies conference on Tuesday in New York City. Here's what a few of them had to say:
JP: You get asked a lot, "When is something going to get into production? When can it be done?" You really have to work at each use case and answer that for each use case.
We've seen examples of supply chain where a real-world example has been done and published and proven out using the technology ... We've also seen cases of trade finance, we've seen cross border settlements.
There's been many use cases — very small and manageable use cases. Where you're managing one agreement or contract end to end, it can be very manageable and you can achieve that sooner rather than later. When you're looking at use cases where you're managing and having to reconcile the balance of thousands and thousands of transactions that are done on a daily basis, it's a little different.
A proof of concept will only get you so far. When you start to look at implementation — how data is managed, how reference data is managed and incorporated, how you leverage smart contracts and the extent to which you leverage smart contracts — there's a number of key design decisions that we've had to make that we weren't necessarily expecting because we'd been through the proof of concept, we thought that we had all of this functionality that's already been built so let's just move it onto a new technology. Sometimes it's not that simple.
AD: It's easy just to say, "Are we there yet?," as though there's some sort of arbitrary milepost that we pass and all the sudden everyone's using a distributed ledger.
But really there's going to be the production parallel pilot, and then there's moving the MVP — the minimum viable product — into production, and then there's expanding the flows, and then there's making sure that we've checked all the boxes around operation and cybersecurity and can this scale?
And then you start driving more adoption and you're bringing on more users, you're onboarding across the market.
And so if you look at it as more of this sort of journey, it's a little bit frustrating to be asked, "Are we in production yet?" We're certainly working on it, but the technology is changing so fast that I think we need to be cautious not to design the systems of tomorrow with today's limitations, whether you're talking about the ability to scale or the transactions per second or the privacy constraints that we have, the technology is changing so quickly that if we move things to production and sink multiple double-digits in millions into really changing market infrastructure, we want to make sure that it's going to withstand the next generation's test of time.
TB: We just announced ... a new channel solution that leverages APIs to basically communicate and orchestrate with a blockchain-enabled platform for accepting private securities and it links to Citi's payment infrastructure on the cross-border side. We've been working on it for about 18 months ... and we've had some limited number of transactions go through.
We talk about minimum viable product, but in this kind of a technology we also discovered that you need to have a minimum viable ecosystem. Because by the fundamental nature of this technology as a multiparty, pretty much arms-length away from each other looking at all of the information in real time and trying to satisfy that it's all accurate and to be able to transact on it. So getting an ecosystem together and working through disparate processes that are mostly in the analog and legacy world, and then being able to transplant them into digital and keep it all in sync across a major ecosystem is, I think, one of the major learnings we had. ... "is it scalable?"
The ability to communicate across these platforms in very high volume transactions with extremely rich messaging embedded in those communication protocols is something that really needs to be tested out. We really need to understand.
VB:These are a series of experiments that are leading us towards production. ... It is a social-technical problem, not just a technical problem. Because in the enterprise itself there are various consequences. The compliance people, the legal people, the technical people, and IT security — all of those have to be martially together in order to solve that problem.
And then once you have solved that problem then you have to talk about how we are going to take it outside because the real benefit of this technology is going to come through association with other banks and mutualizing your source of truth with other organizations. For that purpose, we also joined several consortia like R3 [and] Hyperledger ... So that gives you a platform in which you can safely collaborate with people from other banks and build a solution for a particular problem.
But in the end it has to be through cooperation — through mutualization of your interest and the actual technical details — that you are going to achieve liftoff. And the emerging properties of the system ... are only going to be seen through a small-scale experiment where you are going to say, "OK, now what happens when we hook all of this together?" Those problems are going to emerge when you actually do the work.
So it's going to be a journey, and there's also going to be the whole issue of the platforms that are available are not mature — at least on the enterprise level — so that is the other part of the puzzle that is keeping us from launching completely off the cliff ... so that we don't fall splat on the floor of the canyon.
Suzanne Cluckey / Suzanne’s editorial career has spanned three decades and encompassed all B2B and B2C communications formats. Her award-winning work has appeared in trade and consumer media in the United States and internationally. She is now the editor of ATMmarketplace.com and BlockChainTechNews.com