QCon London 2026: Engineering Challenges in Connectivity and Production-Grade AI Systems

QCon London 2026 marks the conference's 20th anniversary with a program emphasizing engineering rigor in distributed systems and production AI deployment. Scheduled for March 16-18 at London's Queen Elizabeth II Centre, the event highlights two critical tracks addressing systemic challenges in modern software architecture.

The Connectivity Imperative: Beyond Shiny Objects

Daniel Bryant's "Connecting Systems: APIs, Protocols, Observability" track confronts a persistent industry blind spot: the neglect of distributed fundamentals amidst AI hype. As Bryant notes, "All too often, we get excited about the latest technologies when we should focus on API design, cohesion, and coupling." This track prioritizes three foundational areas:

1. Network-Level Engineering: Strategies for managing traffic patterns and failure domains in increasingly dense service meshes
2. Model Context Protocol (MCP): Evaluation of this emerging standard for agent-to-agent communication and its implications for API contracts
3. Production Observability: Implementing actionable telemetry that survives partial system failures

The track's emphasis reflects Bryant's contention that without these fundamentals, AI-driven connection growth leads to unpredictable failure modes.

Operationalizing Non-Deterministic Systems

Hien Luu's "AI Engineering" track tackles the gap between experimental LLMs and production reliability. As enterprises shift from proofs-of-concept to mission-critical deployments, Luu's sessions address:

• Determinism Challenges: Techniques for constraining non-deterministic outputs in enterprise environments
• Agent Reliability Patterns: Architectural approaches for autonomous systems operating within bounded contexts
• Validation Frameworks: Moving beyond accuracy metrics to holistic system performance evaluation

Luu emphasizes that agentic systems introduce novel failure vectors requiring specialized deployment strategies distinct from traditional software.

Architectural Tradeoffs in Practice

Both tracks highlight recurring engineering tensions:

Challenge	Connectivity Track	AI Engineering Track
Abstraction Cost	Protocol overhead vs. flexibility	LLM unpredictability vs. capability
Failure Isolation	Network partition resilience	Agent interaction fault boundaries
Validation	Contract compliance	Non-deterministic output verification

These practitioner-led sessions avoid theoretical discussions in favor of battle-tested implementations, including case studies from finance and transportation sectors.

Complementary Tracks

The full program includes 13 additional tracks spanning:

• Core Engineering: Distributed system debugging, performance optimization
• Emerging Tech: WebAssembly, AI-influenced architecture
• Organizational Impact: Staff+ engineering, AI's effect on team structures

Peer selection ensures sessions focus on actionable insights from production systems rather than hypotheticals. Post-conference workshops include InfoQ's ICSAET certification for emerging technology architects.