Clik here to view.
Platform and site reliability engineering teams will see a big impact on their work from generative AI, predicted panel members at the virtual DevOps Modernization Summit hosted in March by Harness, a software delivery platform.
Developers have been the focus of much of tech industry discussion about GenAI to date. But it’s going to change work throughout engineering organizations, the panel said.
“There’s a lot of buzz around this, which is great,” said Srinivas Bandi, vice president of engineering and product at Harness. “It’s an innovation that has changed the industry for the better, but I also personally feel there’s a lot of downstream impacts as a result of code generation.”
Product and Development: A Tighter Bond
GenAI will lead to tighter, more inclusive relationships — specifically with product management — in the software lifecycle development cycle (SDLC), predicted Wes Whitlam, director of software engineering at Rockwell Automation.
“The introduction of tools to a product management team is ultimately going to help an engineering team and staff as well,” Whitlam said, adding he sees more ability to have visualization and mock-up wireframes at the product management level.
“When that’s introduced to a team at design if we’re talking SDLC, you really have this conduit and exchange of information that is more tangible, that you can edit, update, and then refactor if you need to,” he said. “And even just iterate on, to say, ‘Hey, do we have what we need to go to the next level?’ Which is going to eliminate toil and waste, and keep an engineering team in a flow state, which is what we all really want of our teams.”
Bandi added, “Another way that I can personally also see this improving is, there’s a product spec, and there’s ultimately what’s delivered.”
Another panelist also said the development will work more closely with product managers in the new paradigm. “I just think [of] the creativity of the product manager, their ability and what they can do and their thought of where this can go,” said Nick Colyer, head of digital engineering at AHEAD, an IT services and consulting firm based in Chicago. “It’s going to get interesting to see that tightly coupled.”
Four Areas of Evolution
GenAI will trigger significant evolutions in infrastructure automation and platform engineering, predicted Colyer. AHEAD is already seeing how generative AI is revolutionizing areas of the SDLC both internally and externally with enterprise clients, he said. He compared GenAI’s impact as significant as the shift seen when cloud computing emerged.
“From my point of view, we have teams ramping [up], hiring significantly in the space across our modern apps, data, platform engineering space, and platform engineering, or developer platforms,” he said.
The four areas he predicted will show impact:
- Infrastructure automation and platform engineering.
- The internal developer portal, which he said is probably the number-one request today from development teams. The portal “is probably one of the number-one requests today from development teams,” Colyer said. “Our modern apps teams are working heavily there.”
- The data internal developer portal, because data teams need to provide data services through the platform as well, with support for documentation and knowledge management.
- Developer experience and how to measure that and how productive teams are.
“Get your house in order, platform engineering,” Colyer told audiences at one point in the discussion.
He predicted more of everything — more documentation due to increased productivity, more code reviews and more security scans of components.
“I think those are the concerns of how do the operations side of this, the platform engineering, [to] keep up with the productivity side of things,” he said.
AI Overlays on the SDLC
Data and analytics would become key to managing the increased productivity GenAI will create, said Colyer.
Developer workflow will include GenAI overlays that provide more data and insights into that workflow, according to Whitlam.
“What we’re going to see is the exposure of more data that we can do good things with,” he said. This will give more visibility to developer workflow than DORA metrics or product-portfolio management tools, he added.
“A GenAI overlay on something like that, saying, ‘Here’s where your problem may be, if it’s there’s a constraint or a bottleneck or room for optimization, that’s going to be able to help engineers, engineering leaders, as well, to say, this is a trend that’s going in the wrong direction, we need to go fix something,” he said. “We’re going see a GenAI overlay on just the measures that we feel are important for our community.”
A ‘Bash and Trash’ Approach to AI
Whitlam’s employer, Rockwell Automation, is taking an experimental approach to generative AI, he said.
“We want to create a safety net and an area of experimentation first and foremost, when we talk about DevEx,” he said. “Our engineers are already piloting and using GenAI in their personal lives because they don’t want to miss a boat and they want to get ahead of things.”
Internally, they call their approach “bash and trash,” which means they want developers to be able to build something and throw it away without an outcome, or some purpose or intent, he said. The point is to gain knowledge without punishing experimentation.
There are, Whitlam acknowledged, legal compliance issues to consider. But Rockwell has found opportunities even in creating something as simple as SQL statements but without touching the database by using GenAI, situations that are “trustworthy enough,” he said.
At Harness, the focus is on giving generative AI to the users, and internally on how to leverage GenAI to bring more functionalities to end users.
“In terms of just giving that power of GenAI to our users, we want to do everything in the SDLC, from helping them write code faster, to making sure that builds run faster, that test runs faster, to making sure that they’re writing more secure code, making sure they can adapt new things like [internal developer portals], how do they bring efficiency to their organization? How do they do it with the lowest cost possible?” Bandi said.
The company’s goal, he said, is to “give that boost in terms of developer productivity, give them superpowers, and how to reduce the pain and increase the pleasure of working on all this.”
The post Platform Engineering and GenAI: ‘Get Your House in Order’ appeared first on The New Stack.
Generative AI's impact will affect workflows all throughout the software development life cycle, warned a panel at the DevOps Modernization Summit.