Autonomous operations fundamentally changes the speed, quality and cost of service delivery. This is the same method Google and Amazon have used to upend business after business and it is disruptive. You can start small, but at the end of the day, autonomous operations is nothing less than technology-led business transformation. You need the right platform and the right partner.
Some things to consider:
1) Start small, plan big – You can start with autonomy for one service or a set of services, but the ultimate scope should be full autonomous operations, or as near as you can get to it. Broad thinking should set the requirements and the use cases use as the puzzle pieces to realize the overall plan.
2) Deployment automation vs. Self-driving services – Don’t confuse these two. Deployment automation is important, but all it does is start the car. We are after self-driving services. Services that can navigate the real world on their own. That requires a continuous loop process like OODA: Observe, Orient, Decide and Act. Things change, autonomous ops responds intelligently, they don’t just bail out when something unexpected happens. Instead, the response is unique to the situation and utilizes current telemetry collected across the entire service. Deployment automation doesn’t have that. Self-driving for services requires deployment plus the functions only a continuous loop process gets you: assurance, dynamic SLA, healing (beyond reboot), optimization, maintenance, upgrade, compliance and policy operations.
2) Platform – You need a cross-domain platform for service level autonomous ops. You have to be able to work across all domains and data sources, pulling together all of the expert resource domain controllers that you already own, and operating through and across them to achieve the goals of the services. The same requirement holds true for data. You need the KPIs and KQIs to operate a service, but you might want to factor in insights from your ML platforms or customer app performance data or even the weather. You should be able to design and execute the operations of any service in your organization, based on any data, with the same platform.
3) Designability – Your brand and your approach to the business are unique. They are how you differentiate and compete. The platform you choose for autonomous ops must be able to reflect your approach: to receive your design and make it real. It must be designable by you and your team.
4) Abstraction – Transformation to autonomous ops will take years. During that time all sorts of advances will be made in infrastructure and cloud. Suppliers will come and go, and inter-operator partnerships will become important. You need a powerful abstraction layer as part of your autonomous ops platform to decouple and converge IT resources, data sources, ops and business policy into systematic deployments of services.
5) Progression for your People – Human Capital Transformation (HCT)™ has to go hand in hand with the autonomy transformation. Your personnel should become the designers of operations as the machines become the doers. It’s the right way forward for everyone. This is as critical as any other item on this list. Look for a partner with a proven program for HCT. Nothing less will do.
6) Alignment to the Organization – Organizations have structure and they have culture. The platform and approach you choose for autonomy must respect and fit with your structure. This cannot be a game with just one winner at the end. The transformation to autonomous ops must create a digital reflection of the org as well as the operations.
7) Integration to Data Strategy – Autonomous ops is a form of AI. It may be the most critical form for your business as it is the intelligence that actions the information you have or develop. Some of the information used will be simple KPIs and KQIs. Some will be insights generated by other components of your data strategy including other AI/ML/Analytics tools designed for that very purpose. Integrate the plan for your autonomous ops ‘action platform’ with your data strategy and put both raw data and insights to work.
8) Top down approach – Since the beginning, IT systems have been built from the bottom up. Buy the box, plug it into the wall, load the OS, attach to the network, get an IP, etc. We have reached the limit of the bottom-up approach and it’s time to work top-down. Autonomy starts with goals and works toward the infrastructure. If you don’t start with the goal, you can’t win. If you can’t preserve and action the goal, you can’t win.
9) Modularity and Reuse – Small is OK, but the winners will be the ones who progressively transform the entire organization over time. The key to this is modularity and reuse. You want a platform and an approach that lets your team enroll service after service, resource after resource into a common approach to autonomy. This has to be done in a modular manner, however, so that once you have described autonomous operations of a component once, that module can be reused every time that component occurs in the organization. Modularity, combined with abstraction, radically simplifies both transformation and maintenance.
Sea Street’s platform and services check every box above. We have industry leading experience from our deployments with Fortune™ – caliber customers. We have a lot more to say on all of these topics and stand ready to partner on your transformation.
Ask us. We know the way.