CEO Bill Staples Explains Why Developer Tools Miss 88% of Innovation Cycle—and How Agentic Orchestration Addresses It
Software development tools promise revolutionary productivity gains. AI-powered coding assistants deliver 10x improvements in code generation speed. Yet developers spend only 52 minutes per day actually writing code, meaning even dramatic coding acceleration produces marginal impact on total innovation velocity. GitLab CEO Bill Staples opened GitLab Transcend—the company’s virtual event for technology leaders—by framing this disconnect as the “AI paradox in software delivery,” then demonstrating how the company’s Intelligent Orchestration platform addresses the challenge by automating routine tasks across the entire software lifecycle, not just the coding phase.
The event showcased GitLab’s strategic pivot from DevSecOps platform to what it now positions as an intelligent orchestration platform capable of deploying AI agents across planning, development, testing, security, and deployment workflows. Keynotes, customer sessions with Southwest Airlines, partner discussions with Oracle Cloud Infrastructure, and product demonstrations illustrated how this approach differs from point-solution AI tools that optimize isolated tasks without addressing the coordination overhead, compliance requirements, and workflow dependencies that consume most software team capacity.
Why Coding Productivity Gains Don’t Scale to Delivery Velocity
Staples’ framing of the AI paradox captures a problem many organizations have experienced but struggled to articulate. When developers code 10x faster but coding represents less than 15% of their workday, overall productivity gains remain incremental. The remaining 85% involves activities AI coding assistants don’t touch: requirements clarification, code review coordination, security scan remediation, deployment approvals, incident response, compliance documentation, and cross-team synchronization.
These activities aren’t optional overhead. They’re structural requirements of enterprise software delivery where regulatory compliance, security controls, and operational stability matter as much as feature velocity. Optimizing code generation while leaving these processes manual creates bottlenecks that simply shift from one workflow stage to another.
GitLab’s Intelligent Orchestration strategy addresses this by positioning AI agents as autonomous participants across the software lifecycle rather than task-specific assistants. Chief Product and Marketing Officer Manav Khurana detailed three core components: the Agentic Core (combining GitLab Duo Agent Platform with unified context), Unified DevOps and Security (simplifying end-to-end processes), and Enterprise Guardrails (maintaining control while enabling speed).
“The reality is teams want AI for hundreds of use cases across the software lifecycle, and adding AI feature by feature simply doesn’t scale,” said Khurana. “With GitLab’s platform approach, teams can orchestrate AI agents across planning, development, testing, security, and deployment using the same context, permissions, and security model. That’s how I see AI becoming operational, shareable, and governed across your organization.”
Key Insights at a Glance
- Core productivity insight: Developers spend only 52 minutes daily writing code; 10x coding productivity gains yield incremental total delivery improvements without lifecycle-wide automation
- Platform architecture: Agentic Core combines GitLab Duo Agent Platform with unified context; Unified DevOps and Security simplifies workflows; Enterprise Guardrails maintain governance at scale
- Customer validation: Southwest Airlines using GitLab Duo Agent Platform for mission-critical software delivery requiring 24/7 reliability; adoption examples from Ericsson, Deutsche Telekom, Barclays
- Assessment program launch: New maturity measurement tool launching next month to help organizations chart software delivery modernization paths
- Developer engagement: Virtual hackathon running through March 25, 2026, for custom agent creation; winning projects gain permanent placement in GitLab’s AI Catalog
From Point Solutions to Lifecycle Orchestration
The distinction between AI-powered tools and agentic orchestration platforms matters more as enterprises scale AI adoption. Point solutions optimize specific tasks—code completion, test case generation, security scan prioritization—but operate in isolation. Each tool requires separate authentication, maintains its own context, and lacks visibility into dependencies across the software lifecycle.
Orchestration platforms like GitLab’s approach unify these capabilities under shared context, permissions, and security models. An agent triaging security vulnerabilities can access the same project context, code history, and deployment schedules as agents managing test automation or coordinating releases. This unified context enables agents to make decisions that account for cross-functional dependencies rather than optimizing individual tasks without awareness of downstream impacts.
Southwest Airlines’ adoption illustrates this at operational scale. Airlines operate under regulatory requirements, reliability mandates, and 24/7 uptime expectations that make software delivery risk-averse by necessity. The airline’s technology teams use GitLab Duo Agent Platform to accelerate mission-critical software delivery while maintaining the resilience required for continuous operations—a use case that demands governance, auditability, and controlled automation rather than maximum velocity without guardrails.
Enterprise Guardrails as Competitive Differentiation
GitLab’s emphasis on Enterprise Guardrails reflects a market reality: enterprise software teams need controlled acceleration, not uncontrolled speed. Deployment flexibility that maintains governance, security, and compliance isn’t a constraint—it’s a requirement for regulated industries, public sector organizations, and companies managing critical infrastructure.
Vice President of Customer Experience Sherrod Patching shared results from organizations including Ericsson, Deutsche Telekom, and Barclays, then announced an assessment program launching next month to help organizations measure software delivery maturity and chart modernization paths. This productized assessment approach addresses a common barrier: enterprises struggle to evaluate their current state and identify specific workflow bottlenecks before investing in platform modernization.
The partnership discussion with Oracle Cloud Infrastructure highlighted another enterprise priority: pre-validated, enterprise-ready solutions that reduce deployment risk. “Our managed services program with Data Intensity allows customers to focus on innovation instead of operations by delivering a fully managed GitLab experience on OCI with enterprise-grade security and SLAs,” said Victor Restrepo, group vice president of North America engineering at Oracle Cloud Infrastructure. “Together, GitLab’s intelligent orchestration and OCI’s cloud economics give organizations the flexibility and performance they need to succeed.”
Measuring What Actually Matters
Product demonstrations showcased AI agents collaborating across the software lifecycle with visibility into their impact on engineering metrics—not just task completion rates, but actual delivery velocity, cycle time reduction, and quality improvements. This measurement focus distinguishes platforms that claim AI capabilities from those that can demonstrate business impact.
The virtual hackathon launching through March 25, 2026, invites developers to create custom agents and flows, with winning projects earning permanent placement in GitLab’s AI Catalog. This community-driven approach to agent development acknowledges a reality: enterprises need domain-specific agents addressing workflow variations that generic solutions can’t anticipate.
As software organizations move beyond experimenting with AI coding assistants toward systematic automation across delivery lifecycles, the question isn’t whether AI improves productivity. The question is whether productivity improvements concentrate in activities that actually constrain delivery velocity. GitLab’s bet is that lifecycle orchestration matters more than point-solution optimization—a thesis Southwest Airlines and other enterprise adopters are validating in production environments where reliability can’t be sacrificed for speed.
About GitLab
GitLab is the intelligent orchestration platform for DevSecOps. GitLab enables organizations to increase developer productivity, improve operational efficiency, reduce security and compliance risk, and accelerate digital transformation. More than 50 million registered users and more than 50% of the Fortune 100* trust GitLab to ship better, more secure software faster.
*Fortune 500® is a registered trademark of Fortune Media IP Limited, used under license. Claim based on GitLab data. Fortune 100 refers to the top 20% ranked companies in the 2025 Fortune 500 list, published in June 2025. Fortune and Fortune Media IP Limited are not affiliated with, and do not endorse products or services of GitLab.
