Coming soon: quality-of-service guarantees for decentralized technologies
Why escrow-based service-level agreements (SLAs) are the secret to enterprise adoption of blockchain
During my entire career I have been responsible for delivering, operating, and improving enterprise-grade services — telecommunications, IT, banking — to global customers. I have learned two important lessons when operating the services in more than 65 countries around the world:
- Although the consistent delivery of platform service levels globally is key, the service experience is always local to the country. What matters is the service experience that happens in the real world, not the virtual one that exists on paper in contracts. The real world is where people live their (working) lives and services are consumed.
- Transparency in service reporting, the ability to learn and improve from under / over performance, and a fair penalty / reward system are all essential to a sustainable and successful long-term partnership in the enterprise environment.
For decades, service-level agreements (SLAs) have been central to enterprise outsourcing to IT and network providers, including software-as-a-service (SaaS) solutions. Managed services in particular come in many shapes and forms: from managing the multi-vendor environment to providing service guarantees on a specific technology stack. Managed services governed by one or more SLAs can be unbundled and agnostic to any underlying tech or they can be bundled with a certain tech platform (in part of boost enterprise confidence and reduce risk in adopting that platform). Both of these approaches matter for blockchain.
SLAs are understandably popular, because they provide enterprise customers with guaranteed availability and performance at agreed cost levels, and vendors with a stable revenue stream. This is critical when it comes to widespread adoption of new technologies. Being a good vendor is one thing; being a good customer is just as essential. Both are investing and committing to a longer-term relationship, so it has to be fair and balanced.
Yet the conventional SLA-based procurement model is not without its drawbacks. Contracting cycles often range from months to years, requiring major upfront investment on the part of both customers and vendors into requests for proposals (RFPs), solution engineering, legal, and implementation. Once a contract is in place and services are up and running, customers and vendors frequently face change management processes due to the nature of ever-evolving technologies and business requirements. In the event of contractual disputes, compliance and enforcement also take up significant resources. The biggest challenge I have seen in my career is that often these longer-term contracts can become a burden to staying agile and innovative. With the rapid pace of innovation, it is just not possible to predict how the market will look years from now. At the same time, a provider of the services needs to make a fair return on investment (ROI) too and requires a multi-year lock in.
Fast forward to 2021. More and more enterprises are exploring blockchain and other decentralized digital technologies and their governance mechanisms. Many are willing to actively experiment with open, public networks for pilots and proof-of-concept applications. But when it comes to deploying business-critical systems in production, they tend to default to existing operational practices and habits.
I see many companies turn to the larger consultancies and service providers to step in and provide advice and potentially managed services around the technology of choice. It can be a good way to manage risk — career risk mainly, as the operational risks have not gone away. This is of course also true in the energy sector, where reliability is paramount and system failures can impact millions of customers and result in billions of dollars of damages. So availability and quality of service are essential. How to deliver on these requirements with an open-source decentralized platform is the next big challenge that needs to be solved.
The answer is not to put a third party in between an enterprise customer and the decentralized technology solution to monitor performance and guarantee uptime. Of course this may provide some benefits, but it is an odd way to approach fundamentally new technologies. Indeed, one of many appeals of open-source tech and open architectures is that it reduces or even obviates the need for third-party intermediaries that add cost, process complexity, and opacity to markets.
In this new dawning era of enterprise blockchain, two big questions are emerging that matter for the sector’s future:
- Can decentralized solutions actually deliver enterprise-grade services with high availability and guaranteed levels of service?
- Could SLAs that don’t rely on single vendors or require bilateral contracts be the secret to bridging from limited pilots and proofs of concept to full-blown commercial deployments?
At Energy Web we believe the answer is yes, to both questions, but we need to stretch our thinking into a new space though where we enter into decentralized service operations.
The rise of the escrow-backed decentralized SLA
Conventional SLAs work well because there is a single counterparty to a given contract. That counterparty is known and contracted (and therefore ‘trusted’) to deliver the services. By contrast, public networks like EW-DOS run on a distributed network of nodes, operated by multiple organizations, that collectively run services and maintain data integrity via a common protocol. There is no single vendor with which an enterprise user can enter into an SLA contract, but rather a network of vendors that collectively provide services. So how should we rethink the idea of the traditional SLA? How can we make chief information officers (CIOs) of enterprises, including energy companies, as comfortable relying directly on decentralized technologies as they are with existing IT and network vendors?
The answer lies in a decentralized SLA. And the basis of this new type of SLA — for enterprises and vendors alike — is distributed, multilateral trust that the services will be delivered according to the agreed SLA.
We expect that engendering that trust in a decentralized environment should involve posting collateral (often called staking in blockchain circles). This idea isn’t new; anyone who has ever rented an apartment has likely experienced the concept in the form of a security deposit. In the context of public blockchains, staking keeps actors honest and provides strong assurances that the blockchain holds a shared truth, not a compromised version of the truth (since staked companies are incentivized to follow the rules to maximize their rewards and simultaneously disincentivized from bad behavior, for fear of losing their staked collateral in escrow).
When it comes to the idea of a decentralized SLA, a key mechanism is the staking of crypto tokens into a kind of escrow account. Particularly when native tokens are used (such as EWT), this can be easily done via blockchain-based smart contracts. Staking mechanisms are already being used to guarantee trust at the blockchain level. Why not use the same mechanism to define and enforce the terms that are typically found in SLAs at the services layer? While this is an arguably groundbreaking idea for enterprise IT procurement, it’s already becoming a standard practice for blockchain operational and governance mechanisms.
By extending the concept of token staking beyond the trust layer of blockchains (i.e., the maintenance of the distributed ledger database itself) into the services and application layers of the technology stack, it is possible to provide guaranteed uptime, latency, throughput, and other performance measures that are customarily defined in service levels. This approach has the potential to unlock the full potential of decentralized technologies for enterprise customers and service providers alike.
For enterprise customers: A decentralized SLA allows you to easily procure IT and network services by opting into a pool of services provided by multiple vendors at once, and automatically enforce contractual terms in a trusted way. Not only would this approach reduce procurement costs, but it can also improve overall performance and reliability relative to the status quo thanks to the inherent resilience of a decentralized network. Just as it’s generally better to invest in an index fund instead of picking individual stocks, why bet on specific companies (with vendor lock-in) when you can hedge your bet across the entire market?
For vendors: While in some respects an open-source protocol may commoditize certain IT infrastructure services, it also lowers the barrier for customer acquisition and simplifies service delivery. It’s hard to make a profit on managed services today due to the aforementioned upfront contract costs, fierce competition, and commoditization of IT infrastructure. In a decentralized SLA approach, vendors have much wider reach and fewer liabilities (e.g., you can stop any given service and focus on others, while the overall network keeps going).
Service providers would operate nodes that provision various services to enterprise users. They will get paid for the delivery of the services by the enterprise customers. In return these decentralized service providers would stake tokens against performance guarantees (e.g., uptime, latency) to the enterprises using their services. In the event of performance failures, network and service users impacted by degraded performance or outages would be paid out of the staked tokens through slashing part of the staked tokens.
Bringing the decentralized SLA to the Energy Web tech stack and ecosystem
Our goal is to put this theory into practice with EW-DOS, starting with the Utility Layer services. This approach will deliver all the benefits of conventional SLAs (i.e., pre-defined performance requirements and penalties), while remaining vendor- and technology-neutral, highly flexible, more resilient, and lower cost (e.g., by reducing complexity and point-to-point integrations, automating compliance and enforcement) than conventional bilateral arrangements. The payment by the enterprise customer and the staking of the services providers will be done with native token EWT, which is seamlessly integrated into the open-source technology stack.
For example, consider a grid operator who builds an application that enables distributed energy resources (DERs) to participate in an energy market. The application leverages the power of decentralized identifiers (DIDs) to not only streamline the enrollment of customers and assets, but also manage permissions and coordinate data exchange among market participants. Depending on the size of the market, there might be dozens to hundreds of asset types, and thousands to millions of customers and assets. In this scenario, the application might require multiple Utility Layer services like the Key Manager (for distributed key generation and recovery) and Identity and Access Management Cache Server (for high-performance data structuring and availability for DID verifiable credentials).
Instead of locking themselves into a single vendor or product up front (when there is perhaps uncertainty about uptake), the Utility Layer offers a pay-as-you-go solution that allows the grid operator to procure identical services (in terms of cost, performance, etc.) from multiple vendors at once without having to manage separate bilateral contracts — akin to universal roaming for mobile phones, but for enterprise applications.
A natural skeptic might ask: How is this possible? How do you provide multilateral, guaranteed service levels (in terms of both quantity and quality) without direct engagement and contracts between parties?
I think it is possible, and when it works it will address the two main challenges I referenced at the start of this article. This new type of architecture will enable very transparent reporting on actual service quality and be delivered extremely close to the customer who consumes the service in the real world.
Also remember the famous point that was raised in 1880 by Henry Morton, president of the Steven’s Institute of Technology on the Edison light bulb: “Surely, everyone that is acquainted with the subject will recognize it as a conspicuous failure?”
And so to those who stay skeptical I would say, join the trials and experience it yourself. Expect to hear more from Energy Web on this soon.
Walter Kok is the CEO of Energy Web.
