AWS is Wooing Developers, Yet Again

Back-to-back AI-led product releases by AWS indicate that the company is doubling down on software development. At its DevSphere India 2025 event in Bengaluru, the company introduced a new AI-led approach called the AI-Driven Development Lifecycle (AI-DLC).

Unveiled by Swami Sivasubramanian, VP of agentic AI at AWS, AI-DLC is a methodology that places AI at the core of software development and is now freely accessible on the AWS Builder Center. “Our goal is to help customers build and deploy software faster, with AI managing implementation and humans focusing on strategy,” he said during the keynote address.

In the days leading up to the event, AWS also launched several agentic software development tools such as Kiro and Strands Agents, expanding its portfolio alongside Bedrock, Amazon Q Developer and the AWS CLI.

The company said that AI-DLC is designed to work with these tools and allows teams to condense software development timelines from months to days by using AI to handle tasks like planning, coding, testing and deployment with human oversight at each step.

Instead of just prototypes, AI-DLC starts with real-world business problems and integrates AI into every stage of the software life cycle. At each step, AI proposes a plan, outlines its approach and seeks human validation before execution. “Humans review what it has done and validate the plan, address any issues,” Sivasubramanian said.

Early Customers

AWS shared that Wipro and Persistent Systems are among the early customers of AI-DLC. 

Wipro CTO Sandhya Arun said the IT company partnered with AWS to test the AI-DLC in a real-world setting. She shared that a small cross-functional team from three global locations came together for a 20-hour workshop to simulate the AI-first build approach.

“We said, nothing that AI can do will be done by a human,” she said. The team, which included product owners, developers, testers and business leaders, tackled an enterprise-scale problem with the full complexity of global operations.

By following an AI-first development model, the team delivered multiple models, built a React-based UI and integrated MuleSoft APIs, all within 20 hours. “It was truly enterprise scale in terms of all the enterprise constraints,” Arun said.

Previously, Persistent Systems had used Amazon Q Developer to upgrade legacy Java code, which was 83% faster than doing it manually. Neil Fox, senior vice president at Persistent Systems, said that the future is outcome-based and not based on lines of code.

Fox argued that while AI coding has sparked excitement, organisations still struggle with adoption and measurable benefits. “We need a new manifesto,” he said, comparing the moment to the birth of the Agile Manifesto in 2001.

In parallel, AWS is working to prepare the future workforce for AI-native careers. Students enrolled in AWS Academy now receive 12 months of free access to AWS Skill Builder, a platform offering foundational and specialised AI content. 

For developers, AWS has also launched the AWS AI League, a competition encouraging hands-on experimentation with real-world business problems using generative AI. The programme includes up to $2 million in AWS credits for developers who build and fine-tune AI models through the platform.

Human Oversight Still Critical

Anupam Mishra, director for developer programmes for India and South Asia at AWS, said the company’s goal is not to hand over control to AI, but to maintain human judgment throughout the process. 

Sharing the stage with him, Raja SP, head of developer acceleration at AWS, said the key lies in structured collaboration. “The first step is to verify and regulate the plan. That’s where human oversight comes in.” He added that once the plan is clarified, developers can follow it. 

According to Raja, AI-DLC is not just about individual coding tasks. It is a full-cycle methodology for building production software at scale. “It’s especially good for teams building large, complex systems,” he said, adding that the right context for AI is extremely important. “Each step needs to produce a narrower and narrower context for AI to do the next one well.”

Eliminating Delays in Requirements to Execution

Mishra also criticised the traditional sequential flow from product manager to business analyst to engineer, calling it inefficient and alienating. “This should not be the case. AI is already taking care of generating initial documentation and user stories.”

He described how AWS helps its customers bring together product managers, quality assurance specialists and engineers in the same room. Mishra said the process begins with defining the problem statement, which AI then translates into user stories. The entire team provides oversight, and within hours, a diverse set of tasks is ready to move forward.

This eliminates months of meetings. “Everybody’s aligned, nobody’s waiting for their manager’s sign-off and approval for a specific level of employee…All of those decisions are taken in the room in one go, and the work which takes many months gets done in a couple of hours,” Mishra further said.

The post AWS is Wooing Developers, Yet Again appeared first on Analytics India Magazine.

Scroll to Top