Fireside Chat, Interview, Panel
The New AI Challenge: Trust, Control, and Scale | JetBrains x NVIDIA | RAISE Summit 2026
- Shift in AI Adoption Focus: The current inflection point for AI in enterprises is moving from individual tool usage (literacy) to team-level deployment, where teams own the context, architecture, and full product lifecycle.
- JetBrains' Core Challenge: Individual developer productivity gains do not automatically translate to organizational product-level progress; the next stage requires teams to adopt AI at scale safely while maintaining outcomes.
- Organizational Restructuring Strategy: Enterprises must redesign team assemblies to bring engineers, product managers, and UX designers closer to customer problems, creating a "team of builders" with shared context rather than siloed roles.
- Governance Framework Requirements: Agentic AI must be treated as "managed intelligence" within a system, requiring every agent to be observable, traceable, auditable, and attributed for cost and outcomes while remaining under human accountability.
- Three-Layer Governance Model:
- Individual Layer: Supports fragmented tools (JetBrains and third-party) connected to a central governance system; supports ACP (Agent Communication Protocol) for openness.
- Platform Layer: Provides organizational control over policies, defining allowed tools, models, and agents across all entry points, including third-party and MCP integrations.
- Team Layer: Enables shared agent sessions and environments where team outcomes are visible and shareable across the entire project workflow.
- NVIDIA's "Open Shell" Solution: A secure infrastructure sandbox that defines specific agent access rights, policies, and capabilities within the enterprise environment to prevent uncontrolled AI deployment.
- Risk Mitigation Strategy: The ecosystem must address "shadow AI," cost control, security compliance, and risks like prompt injections by enforcing strict policy enforcement and infrastructure-level observability.
- JetBrains' Competitive Positioning:
- Non-Lock-in Approach: Rejects vendor lock-in by partnering with cloud providers and labs while building an open, unified system connecting software development and infrastructure.
- Core Competency: Leverages 26 years of IDE engineering expertise to focus on quality and product-centric engineering, ensuring that speed does not compromise the integrity of complex "brownfield" projects.
- Role Evolution: Engineers must transition from writing code to defining intent, delegating tasks to agents, and maintaining accountability for software architecture.
- Market Outlook: The industry is transitioning from model training to reasoning and now to agentic inference; the critical challenge is extracting value by implementing these tools to solve real business problems rather than chasing artificial metrics like token counts.
- Future Direction: Success depends on the software orchestration layer that manages team leverage, reporting, and analysis of AI effectiveness, moving beyond raw infrastructure to address the details of team management and workflow integration.