Why we must cede to our robot overlords (to do the dull stuff)

Can I get a show of hands – whose spine shudders at the sound of their own phone ringing? If your hand is up, chances are a component of your role (or role in days gone by… the scarring can be permanent) involves operations. Day or night, it’s that dread associated with wondering “What now?”. A few years back, enterprise started outsourcing the problem of supporting key business systems to 3rd party services, and while this reduced the quantityof calls, it only served to increase the quality – now when the phone rings at 3am, you know things are bad. Real bad.

The problem lies in the human condition – individually, we can only do so much, and en masse we nominalise to lazy and inattentive… it’s far easier to respond than to proact. In fact, our SLAs are written to reward *response* times as opposed to a reduction in potential issues as, to be fair, it’s hard to capture metrics on what could have happened but didn’t. Compounding this is the increasing pressure on operational teams to deal with an ever expanding purview. Once upon a time, a service just had to be available. Now it has to meet user experience metrics, maintain a security posture (in a never-ending game of catch-up), global availability thresholds, multi-channel accessibility requirements etc. etc. All the while pressure is being placed on teams to “reduce their footprint”. It’s unsustainable.

The problem lies in the human condition – individually, we can only do so much, and en masse we nominalise to lazy and inattentive

That is why I’m excited, really excited, possibly more so than any of my colleagues (with the exception of Larry himself) about Oracle’s focus on autonomous services. For those unaware, autonomy in this context is the synthesis of existing platform automation services with advanced AI and Machine Learning algorithms to not just provision services, but to proactively patch, load balance, scale and optimise seamlessly and without any need for human intervention. It’s the IT equivalent of the self-driving car, and it’s going to mean a better quality of service at a much lower cost. Most importantly, it’s going to mean fewer sleepless nights for operational teams, and more attention paid to initiatives that deliver value to the business (and your own resume ;)).

Now, moving to a completely autonomous future is not without its compromises. Control to a degree needs to be ceded, and this is something enterprise IT and business teams have struggled with on two fronts – one is around “trusting the SLAs” as outlined by the service provider, the other is accepting the release cadence and quality of a cloud vendor’s product. On the matter of SLAs, I posit this – when service levels are internally defined, there is little vested interest other than pride to achieve targets. On the other hand, cloud providers (at least reputably sized ones) have tangible financial penalties written into their delivery contracts, and this penalty is amplified by the volume of customers leveraging their service, so as much as you think your internal teams want to hit their SLAs, external parties have far more skin in the game.

When it comes to relinquishing control of how and when upgrades are applied, I think back to a time five years ago when I was having discussions with Exchange Administrators as the wave of email momentum shifted towards Office 365. “Email is business critical” and “We need to regression test our integrations and customisations before any change to the environment!” they cried. The reality is, with limited resources, and a lack of buy in from the business, invariably the rollout of feature, and more importantly, security patches would slip between the cracks and fall (sometimes permanently) into the too hard basket. Fast forward a few years and email is now a commodity service that just works. Why? Because like an iceberg, we realised that running the 5% risk disencumbered us of 95% of the pain – what more, it forced businesses to move away from specialisation and accept the service “as delivered” which resulted in compound savings beyond merely running a managed service. The second you start accepting the good, you realise how fabricated a notion the bad actually was.

While the improved levels of service, proactive scaling, monitoring and patching, and recouped sleep certainly appeal (and as a new father I promise you, the latter is gold), the real kicker for me is the fact that all of this autonomy – the 40 years of experience we’ve poured into to these upcoming services, will be available as highly competitively priced subscription based services via. the Oracle Cloud. Where once a clustered, HA Oracle RAC setup would have had you budgeting 7 figures and 6 months of planning, you can get yourself onto an autonomous DB in minutes at less than the annual cost of an FTE. So all of a sudden, Small to Medium Enterprises no longer have to compromise on their database solution, now they can truly get the best for less.

Where once a clustered, HA Oracle RAC setup would have had you budgeting 7 figures and 6 months of planning, you can get yourself onto an autonomous DB in minutes at less than the annual cost of an FTE.

“Hah” scoff the naysayers, “It’s just another Oracle Sales Consultant regurgitating the script”, and sure, we will go out there and espouse the virtues of just about anything with a big O on it. But in a previous life I headed up the Platform Services team for Australia’s largest online payments gateway, and whilst Oracle constantly came up as the preferred DB solution, we simply couldn’t rationalise it financially on the cashflow of what was ostensibly a fintech startup. Ultimately, we settled on a Galera clustered MariaDB database which, whilst functional, would fall over the second you breathed on it. Not only this, we were constantly tuning poorly formed queries, managing ballooning indexes, patching to maintain PCI-DSS compliance and juggling compute instances to optimise performance. The database consumed the vast majority of our time, and stifled some of our ability to optimise and innovate. What I wouldn’t give to have had an auto-tuning, auto-scaling, auto-patching, robustly-clustered Oracle DB for my OLTP at pennies on the pound!

So where are we at right now? Well, we’ve launched our Autonomous Data Warehouse, and our Autonomous Analytics Cloud is just around the corner. We’ll continue to roll out AI and ML smarts to the remainder of our cloud services in the coming months, so when I say this year is a game-changer for Oracle and cloud services in general, I really believe it, for the robot revolution is here. If you’re ready to hand the boring stuff over to C3PO, claim back a slab of free time, enrich that resume and focus your business spend on innovation rather than frustration, then drop me a line and we can tee up a free trial or, better yet, plan out a POC that’ll prove once and for all that the future is autonomous.

I’d buy that for a dollar!

Disclaimer: The views expressed above are the views of the author and do not necessarily reflect the views of Oracle Corporation.


Author: Tom Walker

Principal Sales Consultant - Oracle Cloud Infrastructure

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: