Artificial intelligence is redefining how software is built, shifting the industry from traditional coding practices to AI-assisted, outcome-driven development. As advanced models begin to execute complex tasks autonomously, the role of engineers is evolving from writing code to supervising, architecting, and validating intelligent systems. Ravi Iyer, Senior Director, Software Engineering at New Relic, shares his perspective with DT on the latest AI developments, the changing nature of engineering roles, and how young professionals can stay ahead in this transformation.
AI Advancements Redefining the Nature of Software Development
Ravi Iyer views Anthropic’s latest AI model as a turning point in how AI interacts with software systems. “Anthropic’s latest release is important because it changes the level at which AI operates. We’ve moved from models that focus on “chatting and writing code” to systems that can actively “operate the terminal.” They are designed to refactor entire repositories, identify vulnerabilities, and execute tasks end-to-end. In many ways, this shifts coding from a manual craft to a more orchestrated service, making it a meaningful transition,” he noted.
He further stated, “The market reaction, however, proved to be significantly challenging for legacy tech companies. Industry leaders weighed in, with Mark Murphy, JP Morgan Analyst, remarking that it feels like an illogical leap to say a plugin will replace every layer of mission-critical software. NVIDIA CEO Jensen Huang similarly argued that AI is not replacing software but is built on and powered by it, calling fears of software’s demise “illogical.””
Addressing the broader market reaction, he acknowledged differing industry viewpoints. “I’m more aligned with the latter view. The role of the software engineer will change and broadly, I see two emerging tracks:
Software engineers working with ML and AI engineers will build large models.
Software engineers using large models and agents will build other software and utilities.
“Operationally, traditional agile and scrum structures will also evolve, with team structures changing. Junior engineers may increasingly handle AI-assisted code generation, build and deployment workflows. Senior engineers will likely focus more on architectural integrity, security, and avoiding code bloat or overengineering,” he said.
Engineering Roles Evolving from Developers to Architects and AI Supervisors
As AI becomes deeply embedded in development workflows, Ravi Iyer highlighted a fundamental shift in engineering responsibilities. “With AI becoming more dominant, we are seeing engineering roles become far more value-driven. AI systems are active participants in code creation. Models like Claude Code can generate substantial volumes of code (recent estimates suggest ~5% of code on GitHub is now written by agents). However, production readiness is still constrained by traditional processes. As a result, the engineer’s role is shifting from “Developer” or “Writer” to “Architect and Editor.”
He outlined several key skills that will define future-ready engineers:
1. Context Engineering (not just prompt writing)
This isn’t just about crafting prompts. It is about embedding the right tribal knowledge, architectural constraints, business logic, and compliance requirements into the system so the output is usable, not just syntactically correct.
2. Architectural Alignment
Engineers should ensure that generated code aligns with modern software architecture principles, design patterns, and organisational standards. AI can produce code quickly, but it does not inherently understand long-term maintainability.
3. Security and Ethics Oversight
With AI-generated patches and autonomous refactoring becoming common, engineers should act as trust auditors. They need to validate outputs for hidden logic flaws, security gaps, and compliance risks, ensuring alignment with internal policies and regulatory standards.
4. Domain Expertise
Knowing what to build is becoming more valuable (especially in sectors with stringent regulations) than knowing how to write every line of syntax.
5. Knowing What Not to Build
Equally important or more important is knowing what not to build. AI agents can overengineer solutions or introduce unnecessary complexity. Engineers must apply judgment to avoid architectural bloat and solution overreach.
6. AI Oversight and Debugging
Engineers will increasingly manage hallucinations, validate outputs, and debug AI-generated code, sometimes using another agent to do so. Oversight becomes just as important as creation.
Advice for Young Engineers: Build with AI, Not Against It
For young engineers entering the workforce, Ravi Iyer offered a clear and practical approach. “You’ve graduated into a marathon that suddenly feels like a sprint. My advice is simple: don’t fight the wave - learn to surf it. First, aim to become an “AI-plus” engineer. Don’t position yourself as just a Java developer or a frontend engineer. Position yourself as a Java developer who can design and architect agentic workflows. The edge now lies in how effectively you can work with AI systems.
Second, build a portfolio of intent. Don’t just upload repositories to GitHub. Show how you used AI to solve complex, multi-layered problems. Demonstrate that you can direct the system, provide the right constraints, validate outputs, and refine results.
At the same time, stay strong on fundamentals. Tools will change, models will improve, but core engineering judgment remains important. You should understand when to use SQLite vs. PostgreSQL vs. a NoSQL database. You should know how to question AI-generated code, identify gaps, and challenge assumptions. Fundamentals are what allow you to supervise AI effectively.
He concluded with a call to action rooted in continuous learning. “And most importantly, don’t wait for universities or curricula to catch up. The best time to plant a tree was 20 years ago. The second best time is now. Your real education in this AI era starts the moment you start experimenting with intent.”
𝐒𝐭𝐚𝐲 𝐢𝐧𝐟𝐨𝐫𝐦𝐞𝐝 𝐰𝐢𝐭𝐡 𝐨𝐮𝐫 𝐥𝐚𝐭𝐞𝐬𝐭 𝐮𝐩𝐝𝐚𝐭𝐞𝐬 𝐛𝐲 𝐣𝐨𝐢𝐧𝐢𝐧𝐠 𝐭𝐡𝐞 WhatsApp Channel now! 👈📲
𝑭𝒐𝒍𝒍𝒐𝒘 𝑶𝒖𝒓 𝑺𝒐𝒄𝒊𝒂𝒍 𝑴𝒆𝒅𝒊𝒂 𝑷𝒂𝒈𝒆𝐬 👉 Facebook, LinkedIn, Twitter, Instagram
