Amazon asks engineers to reconnect with customers as AI reshapes software development

Amazon has asked its software engineers and managers to connect more with their customers and understand their businesses better, acknowledging that some have lost the connection with users as the company has grown.

In a memo, Terk Terkowitz, Amazon’s vice-president and distinguished engineer, Worldwide Amazon Stores, laid out the role guidelines for software development engineers, and said employees who understand their products and businesses can take greater advantage of artificial intelligence (AI) and agentic software development.

“As Amazon has grown, we’ve seen some of our ICs (Individual Contributors) and SDMs (Software Development Managers) become less connected to their customers and business. We want to reverse this trend,” he wrote in the memo.


The message forms part of Amazon’s effort to redefine the developer’s role as AI changes software engineering.

Closer to the business

The company urged ICs at every level to understand their customers and their teams’ product direction more deeply. It said the scope and depth of ownership should increase with seniority.

“For SDMs, we also recommend spending time making sure you’re close to your customers and business, and you’re helping the ICs in your organisation to do the same,” Terkowitz noted.

The guidelines maintain a focus on customer, business and team impact. The company said the pace of change makes it difficult to define AI engineering practices that will remain relevant over time.

“The outcomes that matter aren’t changing, even if the inputs, tools, and techniques are changing faster than ever,” the memo added.

Beyond token counts

The memo also warned managers that tracking AI use through token usage, tool adoption, code changes and reviews could encourage behaviour that doesn’t serve business goals.

“We know that using these metrics could, if used as a target instead of a measure, drive negative outcomes, increasing costs (tokenmaxxing is at odds with frugality), and increasing complexity (measuring LOC is at odds with invent and simplify),” Terkowitz said.

The memo also asked managers to look closely at teams using the least and the most AI; understand what holds back adoption and whether heavy users deliver more.

Leaders must learn AI themselves

Amazon also asked its technical leaders and managers to use AI tools in their own work. They should be able to explain where the tools helped, where they did not and why.

The company has asked engineers to start using AI and keep learning, acknowledging that current tools may not help with every task.

“If your work does not benefit from current AI tooling, that is a finding, not a failing. Document why, share it with your community of peers, bring it to your manager and work to find solutions,” Terkowitz said.

Amazon said leaders who spend time learning AI and give their teams time to do the same have seen large gains. It described its vision as the “Renaissance Developer”, an engineer who combines AI tools with human judgement and knowledge of the business.