Endava's AI Agent Revolution in Software Delivery

AI as the Operating Model

Endava's approach to embedding AI agents into their DavaFlow methodology represents a forward-thinking strategy for accelerating software delivery and transforming enterprise workflows. The emphasis on making AI 'AI-native'—meaning it's the first consideration for problem-solving—is a crucial mindset shift. The article effectively highlights the expansion of AI adoption beyond development into areas like legal, project management, and commercial teams, showcasing its versatility. The practical examples, such as replacing spreadsheets with AI-generated applications for pricing discussions, illustrate tangible benefits and demonstrate how AI can democratize tool creation. The lessons learned, particularly treating AI adoption as a behavior change and encouraging experimentation, are critical for successful organizational transformation. The integration of OpenAI technologies across the entire lifecycle, from requirements gathering to deployment, positions Endava as a leader in practical AI implementation.

However, while the article paints a compelling picture of AI-driven transformation, it could benefit from deeper technical dives into the specifics of agent orchestration and the underlying data infrastructure that supports such complex workflows. The 'how' of integrating different AI models and ensuring data privacy and security within these agentic systems remains somewhat high-level. Furthermore, the long-term implications for the workforce, beyond the expectation of 'AI fluency,' warrant more exploration. While the article mentions reduced manual work, the potential for role evolution and the skills required for managing and collaborating with AI agents could be elaborated upon. The reliance on a single enterprise AI platform (OpenAI) also raises questions about vendor lock-in and the potential for multi-model or hybrid AI strategies in the future, especially as the AI landscape continues to evolve rapidly. The article's focus on the 'future arrived' sentiment is inspiring, but a more nuanced discussion of the challenges in achieving this future at scale, such as data governance, ethical considerations, and the cost of widespread AI adoption, would add further depth.

Key Points

• Endava is redesigning software delivery by making AI agents integral to its DavaFlow methodology.
• AI adoption is framed as a fundamental behavior change, not just a tool rollout.
• OpenAI's ChatGPT Enterprise and Codex are used as the enterprise AI platform.
• AI integration spans the entire DavaFlow lifecycle, from planning to deployment.
• Adoption has extended beyond developers to legal, project management, and commercial teams.
• Key lessons include treating AI as a behavior change, leadership active use, fostering experimentation, and early inclusion of non-technical teams.
• The future of enterprise AI is seen as orchestration, combining models, agents, and workflows into integrated systems.

---

📖 Source: How Endava is redesigning software delivery around AI agents