DevOps could be your bottleneck: Introducing the new LiveOps+DevOps model for efficient operations

The separation between development and operations has evolved through various models. In the 1990s and 2000s, a siloed, control-oriented approach was common, leading to issues like ticket-based communication and unclear responsibilities. The rise of DevOps in the 2010s aimed to address these problems by merging developers and operators into a single team, promising numerous benefits such as faster release cycles and improved quality. And at the start, it did look promising.

However, at the same time, the software, as well as the operating environment for it, changed dramatically. We’ve moved from desktop applications to always-online web applications, from single-user to collaboration, from individual web applications and servers to complex architectures and public cloud environments and cloud devops. Not to forget that the cybersecurity landscape has also become much harder to navigate.

All this got more complicated while DevOps team members were supposed to master it all.

Running a successful DevOps team is certainly possible, and many large-scale technology product companies have mastered that skill. It just takes a lot of people and time to get it right. Consider your own DevOps team. Does it include skilled professionals in at least the following areas?

• Frontend development
• Backend development
• Technical and solution architecture
• Quality Assurance
• Observability
• Platform engineering
• Cyber security
• IT Service Management
• On-call practices, schedules and agreed SLAs

Based on what we are seeing in Futurice’s Managed Services business unit, also known as FutuCare, most companies struggle to meet these requirements. For them, DevOps often fails to deliver on its initial promise.

Our venture to a new way of providing value-based maintenance started with the frustration towards traditional service desks. You know the drill: Send a ticket, wait for a reply, “Have you tried to reboot your computer?” Surely we can do better than this in the 2020s, right?

To address this, we established our own service desk, staffed with cloud experts, platform engineers, and software developers, and named it the LiveOps team. This team is equipped to not only respond to but also resolve most issues on first contact (ie. from the 1st level all the way to the traditional 3rd level). By handling these issues directly, we shield our original development teams, who traditionally serve as third-level support, from unnecessary interruptions. This allows them to focus on solving business-critical domain problems.

After a few iterations of adaptation, our LiveOps team is exceeding our expectations:

• 83 % of issues are being resolved on first contact
• 100 % of reaction time SLAs kept
• 93 % of resolving time SLAs kept (with average resolving time of 2 hours)
• Customer satisfaction is extremely high with a 4.9/5 CSAT score

Both our customers and employees love the new system, proving that our approach is making a significant impact.

A DevOps team member is quite often expected to be a jack of all trades while truly being a master of none. Due to their predominantly developer backgrounds, DevOps teams tend to skew towards being 90% Dev and only 10% Ops.

In client organizations, development programs are typically driven by business units aiming for a competitive edge through new software solutions. This creates a scenario where the Product Owner prioritizes new feature development over maintenance and operational tasks.

As a result, operations-related issues are frequently neglected, leading to major, hard-to-fix problems when the software is in or nearing production. To mitigate these challenges, we decided to start to integrate our LiveOps team members into development projects from the early stages. This proactive approach prevents real-life problems from arising, and the results have been remarkable.

Our LiveOps+DevOps working model is based on the principle that LiveOps is not merely a support function.

Benefits

• Best practices and sane defaults for quicker time-to-market: There’s a difference in experience whether you’re doing something for the first or second time, or for the 20th time. Whether it's Platform Engineering, Cyber Security, or Observability, our LiveOps experts bring extensive knowledge to the table. By leveraging this expertise from the outset, we ensure implementations that don’t require last-minute fixes before going into production.
• Experts available when needed. Our whole Modern Managed Services approach is built on the principle that you purchase a service, not specific individuals. If you need a Platform Engineer to create an efficient deployment pipeline for your application, our LiveOps Platform Engineers will be available and can handle it. This approach eliminates the need for you to hire a full-time expert for your development team.
• Great observability leads to better quality. Our LiveOps team employs a robust set of leading monitoring tools and practices. We integrate these into your solution from the beginning and advise your development team on best practices for cloud runtimes, logging, and application performance monitoring. This shift in the development mindset from "Will this work?" to "How do we know this will work?" and "Is the system working well enough?" significantly enhances your application's quality and simplifies tracking KPIs.
• Hidden Ops work gets revealed and addressed. A common question from prospective clients is why their development activities are slowing over time. This is usually because operating a production system is demanding, and developers often underestimate the work involved. By tackling recurring issues—such as vulnerability scanning and dependency updates—with automation and processes like those used by our LiveOps team, we prevent the accumulation of blocking maintenance tasks.
• AI opportunities become easier to identify and realize. While integrating AI tools and capabilities into development workflows is often discussed, implementation can be challenging. However, since the LiveOps team regularly addresses similar issues, it's more feasible to develop AI-driven processes and automations that enhance operational efficiency.
• More time to work on business problems. When the majority of maintenance and operations-related tasks have been moved to the LiveOps team, the DevOps team has more time to deliver business value through new feature development.
• Seamless transition to production. LiveOps+DevOps operating model ensures some of the most uneventful and boring production deployments, by design. We prioritize stability over excitement, resulting in smooth, reliable production launches.
• Smooth operation in production. By incorporating LiveOps early in the development program, we build a solid foundation for handling operational issues. Our LiveOps team transitions from enabling a solution to maintaining it. It's like flipping a switch in our operations room—suddenly, you're supported by our LiveOps Service Desk, backed by guaranteed SLAs.

In this model, LiveOps is not intended to replace DevOps but to enhance it. While DevOps is an excellent methodology for building software, offloading certain tasks to a centralized LiveOps team is both logical and beneficial, simplifying processes and improving efficiency for everyone involved.

A LiveOps team member participates in the DevOps team’s agile meetings and operates as a normal team member. This integration is led by a single Service Lead, who acts as the single point of contact for the client. This ensures smooth development, efficient operations, and a seamless production environment.

More than anything, the LiveOps+DevOps working model offers a practical solution to a problem that, in theory, shouldn’t exist but, in practice, hinders many companies.

At FutuCare, we truly believe that the LiveOps+DevOps model can supercharge most development programs. We’re eager to share our insights on how to implement it effectively.

Feel free to follow our four-step process, which we use to determine the best way to incorporate LiveOps+DevOps to meet our clients’ needs:

1. Assess the current state. Evaluate the number of applications in development and those already in production. Identify which production applications are business-critical and determine their SLA requirements.
2. Determine the potential. Identify common tasks across DevOps teams that could be centralized. Assess how much time individual DevOps teams currently spend on these tasks. Determine if this time could be better utilized or if the tasks could be executed with higher quality through centralization.
3. Build and roll out the LiveOps team in a gradual manner. Avoid moving all Ops work to LiveOps at once. For long-established DevOps teams, accumulated hidden knowledge can hinder this process. Start small by selecting specific work areas, such as the maintenance of deployment pipelines, to transition to the LiveOps team.
4. Let the LiveOps team improve things. The greatest benefits arise when LiveOps not only maintains but also enhances the assets it operates. LiveOps team members, motivated to streamline their work, will drive improvements in shared components, automation, and processes. Ensure they have sufficient bandwidth to innovate without being overwhelmed by their workload.