SAP’s Bharat Sandhu on Gen AI Coding in the Enterprise

Generative AI is no longer just a buzzword—it’s a real, transformative force. In this in-depth interview, Bharat Sandhu, Senior Vice President and Chief Marketing Officer for SAP Business Technology Platform, offers a grounded and insightful look into how generative AI is being used to build enterprise software—and not just prototypes.

From redefining the role of the developer to establishing new protocols like MCP (Multimodal Context Protocol), Sandhu outlines the essential changes companies must adopt to remain competitive. He also shares a compelling vision of where software development is headed, including a more intelligent UI, integrated DevOps, and the rise of autonomous agentic systems.

Companies are using GenAI to boost developer productivity, tackle backlogs, and experiment with new prototypes. While “vibe coding”—voice-based code generation—is still limited to prototyping, professional developers are using AI tools today to accelerate real production work.

Context Is Key for Enterprise AI Success

GenAI models need enterprise-specific context to be truly effective. Protocols like the Multimodal Context Protocol (MCP) offer a framework for safely and intelligently integrating enterprise data with generative models.

GenAI Must Fit into Existing DevOps Pipelines

AI-generated code must still follow the same testing, compliance, and deployment procedures as traditionally written code. The best approach integrates GenAI into the existing software development lifecycle — not as a separate track, but as a productivity-boosting enhancement.

The Future Will Bring Smarter Interfaces and Autonomous DevOps

New user interfaces will abstract away from pure code and allow teams—from business leaders to developers—to collaborate across layers of abstraction. Over time, intelligent agents may autonomously test, iterate, and deploy software based on real-world usage patterns.

Key Quotes: “That’s how we break the old limitations…”

Ideas Over Code

“Writing code is no longer the differentiator—the ideas are. The real value lies in the new insights, the creativity, and the ability to scale those ideas into production. That’s where you still need software engineering depth—systems thinking, architecture—things that go far beyond just code generation.”

Time to Get On Board

“If you’re not already using generative AI for software development, you’re behind. Start with your most senior developers—they know the systems and the codebase best. Then onboard junior developers with guardrails, using GenAI to accelerate onboarding and reduce dependency on senior team members.”

The Importance of Context

“The biggest thing that will prevent companies from scaling GenAI effectively is context. These models don’t know your customer data, your invoice structure, or your application ecosystem. That’s why we’re championing the Multimodal Context Protocol—a smart, managed layer that lets AI safely access enterprise data.”

Evolving UI Development

“We’ll see a major evolution in the UI of software development. It won’t be just code editors anymore—you’ll have layers of abstraction, where business users, product teams, and engineers can collaborate seamlessly. You’ll be able to click from high-level intent down to code and back up again. That’s how we break the old limitations of low-code tools and empower real innovation.”