CloudBolt Lightning Strike 2024: Why automation has failed FinOps, so far

CloudBolt spent some serious time detailing its platform updates, product roadmap, future FinOps vision and wider strategy at its annual gathering of financially focused cloud practitioners this January – and the Computer Weekly Developer Network team was there to cash in (sorry) on the opportunity – this was CloudBolt Lightning Strike 2024.

To remind ourselves, CloudBolt Software is known for its FinOps platform skills and its technology that works to manage the automating, optimising and governing hybrid-cloud multi-tool environments for enterprises.

Kicking off the event this year, CloudBolt chief revenue officer (CRO) Pritish Upadhyay took the stage to rouse the audience and set the scene by welcoming attendees from the different world regions gathered. Clearly a C-suite executive who takes a very employee-facing role, Upadhyay urged attendees to mix & mingle but also to ask as many questions as possible and engage with breadth.

CRO Upadhyay welcomed and gave way to the company CEO.

Craig Hinkley, CEO of CloudBolt Software is previously on the record stating that ‘rarely in the history of business has there ever been a workplace practice that has emerged and been adopted as rapidly as FinOps’ – a statement that he tempered by actually calling this workplace practice a discipline, perhaps alluding to how seriously we should consider FinOps in the context of modern cloud.

Cloud has an ROI problem

Both Upadhyay and Hinkley enthuse on the amount of change that is now happening in this space, “It’s all about using a different vocabulary when speaking to practitioners and users at all levels… we are now at a point where we should be talking about ‘augmented FinOps’ today,” echoed the pair, speaking on stage this January 2024.

An affable Aussie with a naturally upbeat tone, Hinkley suggests that – in many enterprises today – cloud deployments have a return on investment (ROI) problem. He notes that the DNA of the company itself has focused on retooling its customer-centric stance to elevate CloubBolt’s ability to deliver execution and repeatability across its platform for greater functionality and an altogether more flexible and adept approach to FinOps.

It’s all very well to talk about having a new website and a new customer portal (and CloudBolt has brought those fundamentals forward) as it works to move its technology forward, but Hinkley talks of a new drive where the company is set to challenge both the market and the customer base.

Moving up to augmented FinOps

CEO Hinkley talks about the complexity that already exists across modern cloud environments with all the interconnections that pervade within a typical hybrid cloud deployment.

“Engineering teams will be thinking about which cloud, which Site Reliability Engineering (SRE) benchmarks they need to hit, which workload to put where in the multi-cloud hybrid world, which cloud needs to be de-provisioned and retired in a secure way and so much more, the case for augmented FinOps that has AI/ML-informed intelligence functions at its heart has never been clearer,” said Hinkley. “This moving beyond ‘basic’ FinOps to be able to operate across the full FinOps lifecycle with both retrospective and predictive forecasting in an orchestrated automated way.”

In addition to the challenges highlighted here, if we combine the huge number of network connections between cloud instances, applications and data services, we start to realise just how complex the FinOps lifecycle can be. We can also consider the complexities of cost-related cloud actions including everything from the use of ‘reserved instances’ (cloud service provider deals designed to package service levels into cost-saving deals) and then onward to API-first connections designed to accelerate existing customer cloud functions with optimal savings, there is so much to consider.

Urging attendees to get the most out of the rest of CloudBolt Lightening 2024, Hinkley may be a rare breed as an Autralian-born tech CEO, but his vision appears to align with the way cloud operations should now be evolving in the age of AI and augmented, accelerated, automated everything.

Embrace platform engineering

The wider trends here connect directly to the need to embrace platform engineering i.e. focused operationalised infrastructure services spanning serverless computing, kernel-based provisioning and programming and all lower substrate elements of any given computing system.

As TechTarget reminds us here, “Platform engineering is a specialized discipline within software development that focuses on designing, building, maintaining and improving the toolchains and workflows software developers use. Platform engineering provides comprehensive and consistent tools and processes, which enables developers to focus on software development instead of managing underlying toolchains.”

We can also say that platform engineering is an approach designed to bring standardisation to developer workflows and toolchains to create a more consistent and predictable experience within any given software system – it ultimately can result in the creation of internal developer platforms featuring customised, reusable and orchestrated code components which speed up programming functions while also making them more secure.

CloudBolt’s chief product & technology officer (CPTO) Kyle Campos has said that CIOs will ‘institutionalize’ platform engineering and so establish teams to enable these management processes. Campos talks of the need to move away from FinOps silos that do nothing but create friction through unrealized optimization promises. He says it’s all about the shift from ‘motivation’ to ‘facilitation’ and that FinOps practices will become a native part of the journey to ‘golden paths’ for IT systems management.

CloudBolt CEO Hinkley: Move beyond ‘basic’ FinOps into the full FinOps lifecycle with both retrospective & predictive forecasting in an orchestrated automated way.

Where we come full circle in this story is when we realise that (as our title here suggests) FinOps practices – at least up until now, if we buy the AI/ML augmented FinOps concept that CloudBolt is coming to market with – have been failed by existing approaches to automation.

But why is this so?

It largely comes down to FinOps being too rigid for the realities of business. We now stand at a point where there is a critical role for advanced automation in revolutionising the FinOps landscape.

“This whole process begins by addressing the prevalent challenges faced by FinOps teams, primarily revolving around an endless cycle of reactive tasks such as data preparation, report building and data analysis. There is a widespread belief among FinOps leaders that automation is the key to liberating their teams from these manual, time-consuming tasks. Automation before now has been too rigid to match the complex operational needs or real organisations using modern cloud,” explained CloudBolt CPTO Campos.

Eradicating underutilisation & inefficiency

Campus says that when looking at the current state of automation in FinOps tools, there is an underutilisation and inefficiency in enterprise environments. He says that previous tools have shown their limitations, particularly their lack of flexibility and adaptability in handling complex, real-life scenarios, which often leads to increased risks and inefficiencies.

“We are proposing a path forward for FinOps automation,” said Campos. “Organisations need to more flexible, capable and scalable automation solutions that can deliver tangible benefits without adding to the maintenance burden. Automation should be the solution for FinOps to transcend operational noise.”

Looking to the (immediate) FinOps future, Campos joins Hinkley and Upadhyay in the proposition that the application of advanced automation to not only streamline operations but also enable FinOps teams to focus on strategic initiatives that drive business value.

FinOps is not quite up there with DevOps, SecOps and the everything XOps selection pack of Ops connection points, but it is in the ascendancy and worth tracking very closely indeed.

Data Center
Data Management