Developers don’t want to waste huge amounts of time on repetitive, thoughtless tasks, and GenAI helps them get out of that situation.
Specifically, at Amazon, developers previously reported spending 70% of their time on tedious, repetitive tasks instead of coding, which is why they are leveraging new tools like Amazon Q Developer, an AI-powered software development assistant, to reduce monotonous tasks.
The tool helps developers be more productive and write code more securely, while also helping novice developers ask more questions and learn examples from 24/7 assistance.
The proof is in the track record. Amazon CEO Andy Jassy revealed during the company’s Q2 2024 earnings call that using Q Developer’s code conversion agent feature, Amazon migrated 30,000 production applications from Java 8 or 11 to Java 17, saving over 4,500 years of work and $260 million annually in performance improvements. It has been generally available since April 2023. Now, a little over a year later, the impact it has had on developers is easier to understand.
So how exactly does it work and how does it improve your workflow?
Amazon Q Developer Capabilities
Amazon Q Developer generates highly accurate code and enables you to have conversations about that code, including filtering out potentially biased or unfair code suggestions, implementing new code generated from developer requests, and debugging and troubleshooting.
“A big part of this is that so much of what developers do every day isn’t very exciting,” says Doug Seven, general manager and director of AI Developer Experience at Amazon Web Services. “It’s a lot of mundane work. It’s necessary work, but it’s not cognitively very interesting or challenging. If AI can do some of that for development teams, freeing up their cognitive resources to do things that are more interesting, novel, and have more business value, that’s great.”
Early indications suggest that Amazon Q can help organizations increase employee productivity by over 80%. When Amazon Web Services ran a productivity challenge, developers who used Amazon Q Developer were 27% more likely to complete tasks successfully. This is in part because it’s easy to use and requires no training to understand.
Seven likens it to having another developer sitting next to you watching your work and making suggestions — he says it’s like super autocomplete — but also allowing the developer to ask questions about where there might be an issue with the code.
Another way to use Amazon Q Developer is to use AI agents. For example, transpiling code from Java 8 or 11 to Java 17 is tedious and can take up to two days per program. Instead of developers doing this themselves, they can assign that task to Amazon Q.
“In a sense, it can be used like another engineer on your team, who can do the work and then go back for peer review to make sure the work is good,” Seven said.
And once developers truly begin to understand what tasks are best delegated to AI, it can lead to real success.
“Now I would say any solo developer can become a team of developers,” Seven says. “They can assign their work to different agents and do other things while the agents are doing their work.”
Ripple effects
With such a huge shift in productivity thanks to these new tools, one might ask, what job is left for developers? This is a big question with the boom of AI, but most argue that AI will enable workers to do more fulfilling and important work. Additionally, humans will need to stay on top of things.
“Having an AI assistant write code for you is not that different from having it translate an English sentence into French with the correct semantics,” says Jason Andersen, vice president and principal analyst for application development and platforms at Moore Insights & Strategy. “In both cases, the requester needs to understand the situation and provide context for the task.”
There are other gaps developers need to consider when leveraging AI: For example, Andersen said that today’s AI isn’t smart enough to fully understand the nature of applications that consist of thousands of files spread across many systems and locations.
“Given the complexity, an AI agent can’t truly understand the context of previous decisions, potential decisions and constraints,” Andersen says, “so a developer needs to get involved and review the AI’s recommendations to determine whether they’re useful.”
While Amazon Q Developer has guardrails to ensure that AI agents are behaving responsibly, the results you get from the AI may not always be entirely correct, so again, human oversight is required to make sure things are running smoothly.
“I have to say that the developer role has been evolving for a long time,” Andersen says, “The range of skills, including collaboration and coordination, is becoming increasingly sophisticated. So AI not only needs to keep up with what developers do today, but also evolve with the future evolution of the role. I think this is in contrast to other jobs that may one day be replaced by AI.”
