Who You Need on Your Automation Dream Team

Organizations can reduce costs and increase agility by deploying automation in their businesses to handle repetitive tasks and processes.

“The ideal automation dream team combines a good mix of technical expertise with business acumen so they can transform manual processes into efficient, scalable systems,” says Prashanth Ram, cofounder & CTO of Smoothstack, an IT services provider specializing in workforce development.

“When building an effective IT automation team, I look for a blend of technical expertise and mindset attributes that go beyond just coding skills,” says Corey Ercanbrack, chief product technology officer for Vasion, an orchestrated automation company. “Having led engineering teams at companies like Intel, LANDesk, and now Vasion, I’ve found that the right team composition makes all the difference.”

To unlock automation-driven savings, you need a team that understands systems thinking, scripting, integration and orchestration. Systems thinking involves “possessing the ability to see how components interact across an organization,” Ram says. It also involves being able to see the big picture rather than focusing on individual technologies as well as predicting the “cascading effects of changes as well as building holistic solutions as part of an ecosystem, Ram says.

Ercanbrack sees systems thinking as a foundational requirement for automation professionals.

“I need people who can zoom out and see the entire workflow ecosystem rather than just discrete tasks,” Ercanbrack says. “These individuals understand dependencies, can predict the ripple effects of changes and naturally identify bottlenecks that need automation.”

Being able to analyze business processes is a key requirement for joining an automation dream team, according to Roberto Montero, director of engineering at Andela, an AI-powered marketplace for remote tech talent.

Then automation professionals should be able to design and develop bots, as well as integrate them into applications such as enterprise resource planning (ERP) or customer relationship management (CRM) software, he says.

“For scripting capabilities, I look for engineers who can write clean, maintainable code across multiple platforms,” Ercanbrack says. “With AI tools now augmenting this process, engineers must understand both how to write code and how to effectively leverage AI-assisted development to accelerate their work without compromising quality.”

Orchestration, or coordinating multiple IT automation tasks or processes, is another key part of an automation expert’s responsibilities. “Orchestration maturity” is what people need on an automation dream team, and that involves coordinating multiple automated processes with a knack for timing and understanding dependencies and exception handling, according to Ercanbrack.

“The best automation engineers can design workflows that incorporate both traditional logic and newer AI-driven decision points where appropriate,” Ercanbrack says.

Here’s a look at the top roles you need on your automation dream team.

DevOps Engineers

DevOps engineers integrate both development and operations teams across an application life cycle. They deploy a software development practice called continuous integration/continuous delivery (CI/CD) and infrastructure as code (IaC), which involves managing and provisioning IT infrastructure using code rather than manual configurations.

IaC code solutions can “decrease provisioning time from weeks down to as little as minutes while also eliminating the human error aspect and ensuring a consistency within environments across development, testing and production,” Ram says.

At Smoothstack for example, automation through CI/CD improves an organization’s ability to respond to customer requirements, according to Ram.

“Our CI/CD pipelines speed up software delivery cycles, making it possible for organizations to respond even more rapidly to customer needs and market opportunities while maintaining necessarily high-quality standards,” Ram says.

Hire DevOps engineers with experience with monitoring tools as well as containerization systems such as Docker and Kubernetes, Ram advises.

Also look for DevOps engineers with an understanding of TerraForm, a tool that lets IT leaders manage cloud infrastructure using code, advises Vishal Gupta, former executive at Lexmark.

DevOps engineers not only need to know about IaC to automate infrastructure management, but also infrastructure from code, which consists of automatically generating scripts for setting up infrastructure environments.

“IAC is providing the first level of automation, and IFC is providing an even higher level of automation,” Gupta says.

IFC accelerates TerraForm script generation from one day to around two hours, Gupta says.

“But since it’s automatically generated, we still want a human to review it,” Gupta says. “In this case, it’s the DevOps engineer reviewing the script to make sure the machine, the AI, did not miss anything.”

Hire DevOps engineers who understand immutable technologies versus patching technologies, Ercanbrack advises.

“Do they understand what it means to be immutable, what it means to be highly available, how you might build that in whatever cloud offering you’re going after, whether it’s SaaS, whether it’s Azure, or maybe it’s even multicloud,” Ercanbrack said.

“I fundamentally think that any engineer you hire, whether it’s a DevOps or a developer engineer, has to have DevOps as a mindset, which is really about that site reliability, like, how do I scale that out? How do I make sure it’s up and running across tens of thousands of customers at all times,” Ercanbrack said. “And so we have sets of tests that get executed along that way that get pushed out.”

Site Reliability Engineers (SREs)

These IT professionals ensure systems are resilient and self-healing by using software and automation to solve problems. Google originated the concept, Gupta notes.

“Right now, it’s very widespread in the industry, and as the name sounds with SRE, it’s all about hitting a higher level of uptime for the site,” Gupta explains. “So you’re trying to make sure the site is up or the application is up.”

A site could go down if they are short on storage or memory or if they have a misconfiguration, Gupta says.

At Lexmark, SREs examine applications to ensure they can commit to a reliability time.

“The objective for site reliability engineers is to really automate, to come up with the whole methodology for ensuring these things are highly reliable and highly available, and be able to publish these so the customers don’t even have to come to us,” Gupta says.

Lexmark looks for SREs with experience with not just websites but also software applications, Gupta advises. He also suggests evaluating the type of service-level agreements that SREs have implemented as well as the ability to perform site reliability engineering across regions.

Engineers hired to build self-healing networks should have a definitive road map and a clear set of steps, Montero says.

Tech recruiters should look for SREs with knowledge of distributed systems, monitoring and incident response, according to Ram. They should also have programming skills and experience with observability tools.

Because SREs may deal with crises such as outages, they should have solid problem-solving skills to maintain system stability, Ram adds.

In addition to automating, SREs pursue strategies that involve redundancy, chaos engineering, observability, service-level objectives and graceful degradation, which can mantain a system’s core functionality and prioritize partial availability over full outages, Ram says.

Robotic Process Automation (RPA) Developers

RPA developers build bots to automate back-office workflows and repetitive tasks. The bots can streamline business processes. Organizations such as Udacity offer courses in RPA.

“RPA developers follow a very structured approach when they are going to be building bots to automate back-office workflows,” Ram says.

The process of building bots as an RPA developer includes process discovery and analysis, which involves “starting with identifying repetitive manual tasks that are suitable for automation by observing users in action and mapping the current workflows as they happen,” Ram says.

Also hire RPA developers with expertise in solution design. That involves “creation of process documentation that is very detailed, which includes marking down any decision points, exceptions and outcomes that are expected to happen as a result,” Ram says.

In addition, RPA developers should also know about testing and validating of bots.

“Verify the bot is functioning correctly in different environment scenarios that are of a controlled nature before any actual deployment happens,” Ram says.

Hire RPA engineers with proficiency in RPA platforms like UiPath, Automation Anywhere and Blue Prism, Ram suggests. RPA developers should also have experience with business process analysis as well as workflow automation, he adds.

In addition to UIPath and Automation Anywhere, Gupta advises that recruiters hire RPA engineers with knowledge of Microsoft Power Automate. They should also understand the domain and create low-code RPA solutions, which doesn’t require deep software developer knowledge, Gupta says.

Also look for RPA engineers with solid attention to detail. “This is critical for creating robust, error-handling automations,” Ram says.

AI/MLOps Engineers

AI/machine learning operations (MLOps) engineers optimize AI pipelines and automation frameworks. They include data scientists, DevOps engineers and IT and streamline the process of taking ML models to production.

AI/MLOps engineers build remediation scripts as part of observability and diagnostics to analyze why systems go down, Gupta says.

“We typically will leverage AI/ML engineers to figure out the ‘why’ in a faster way,” Gupta explains.

He says these engineers use AI/ML for figuring out observability dynamics or diagnostics. AI/ML engineers also automate remediation to determine steps to bring a system back up.

Hire AI/MLOps engineers that know about infrastructure optimization, Ram advises. This involves “implementing containerization (Docker) and orchestration (Kubernetes) for a consistent deployment experience, putting to use cloud-native services with auto-scaling to handle variable workloads and selecting the appropriate compute resources (CPU/GPU/TPU) based on the model requirements,” Ram says.

Because RPA is an older technology, Lexmark hires more AI/ML engineers, which requires more significant coding and AI skills, Gupta says.

AI/MLOps engineers should also know about pipeline efficiency, monitoring and observability, version control and resource management, Ram says.

Ram says resource management entails “applying parallel processing when and where its applicable, implementing model compression and quantization techniques, and using deployment strategies that are hybrid in nature (cloud/edge) for latency-sensitive type applications.”

Going forward, Ercanbrack expects to see a blending of automation and AI engineers. He says the ability to use AI with an automation context will be key.

“At Vasion, we’ve started looking for what we call ‘AI-native automation engineers’ — professionals who naturally think in terms of human-AI collaboration when designing automation solutions,” Ercanbrack shared.