Technology
QCon London 2027 Shapes Programme Around AI's Impact on Architecture and Team Leadership
The conference's 15 tracks examine how senior engineers can adopt AI practices thoughtfully by understanding the trade-offs and organizational conditions behind emerging approaches.
InfoQ
QCon London 2027 is structuring its conference agenda around a central premise: AI fundamentally reshapes how teams approach architecture, security, delivery, and organization—but no single playbook applies universally. Rather than presenting best practices as universal templates, the conference invites senior engineers to examine the reasoning, compromises, and context that made specific approaches work.
The conference, running April 13–16 at London's QEII Centre, features 15 tracks organized by theme, with more than 75 practitioner speakers sharing details of production systems they have built, the trade-offs they navigated, and what they discovered when designs collided with real workloads and organizational constraints.
Controlling Autonomous Systems in Production#
Moving AI agents from experimental prototypes into live environments demands new safeguarding approaches. Teams must establish behavioral boundaries, validate changing outputs against shifting requirements, and develop incident response protocols for unpredictable autonomous decisions.
Two dedicated tracks address this challenge. "Eval Platforms & Guardrails for Agentic AI Systems" focuses on the evaluation frameworks, feedback mechanisms, and safety controls being deployed around AI-powered applications. "Protecting Systems from Humans and Agents" shifts attention to how security strategy must evolve when failures stem from both human mistakes and autonomous systems.
The underlying goal is not to prescribe a universal guardrail design, but to help engineers understand which control mechanisms address specific risks, where trade-offs exist, and which security concerns merit different placement—whether at the model, application, platform, or operational level.
Rethinking Architecture and Development Workflow#
Integrating AI into existing systems introduces new layers of complexity: novel dependencies, non-deterministic behavior, data requirements, and performance profiles. The architectural question extends beyond simply choosing where to invoke a model—teams must weave intelligent components into systems while preserving reliability, operability, security, and the freedom to iterate.
Three tracks explore these tensions. "Architecture in the Age of AI" examines patterns for embedding AI capabilities while sustaining scalable and dependable systems. moves beyond template diagrams to reveal the real-world compromises that surface when novel ideas encounter production realities.
