Platform engineering is evolving from a tactical extension of DevOps into a comprehensive framework that integrates infrastructure, APIs, data pipelines, and AI agents, setting the stage for enterprise acceleration as we approach 2026. As indicated by the 2025 DORA report, nearly 90% of enterprises are adopting this approach, while teams grapple with the challenges posed by agentic AI, particularly in governance and interoperability. To address these trends, DevOpsCon will host its Platform Engineering Week from May 11–15 at Park Plaza Victoria in London, uniting DevOpsCon, API Conference, and MLcon to reflect these convergences.
“Platform Engineering didn’t suddenly become something else. It quietly became more than we initially planned for,” remarked Sebastian Meyen, head of content at Software & Support Media, as noted by WebProNews. This transformation is designed to overcome the scaling challenges faced by DevOps, including tool sprawl, inconsistent standards, and developer burnout amid increasing complexity. Gartner predicts that 80% of large engineering organizations will have platform teams by the end of this year, an increase from 45% in 2022, which will facilitate the delivery of reusable services across the stack.
The discipline of platform engineering is built on 12 foundational pillars—spanning infrastructure automation, CI/CD management, security, observability, and API governance—that now extend to AI model deployment, treating agents as integral components complete with role-based access control (RBAC), quotas, and policies. Notably, at Spotify, AI agents have generated over 1,500 merged pull requests, reducing migration times by 60% to 90%, according to The New Stack.
As the landscape shifts from DevOps silos to unified platforms, the initial promise of DevOps in breaking down development and operations barriers has been challenged by the reality of scale, resulting in developers facing YAML sprawl and an array of custom scripts. In response, platform engineering is introducing internal developer platforms (IDPs) that abstract away much of this toil, offering self-service “golden paths” that include preconfigured workflows with built-in security, observability, and AI defaults. Spotify’s Backstage project, which holds an 89% market share, underpins over 3,400 users, including LinkedIn and Vodafone, though DIY efforts often stall at just 10% adoption after six to twelve months of maintenance difficulties, according to Roadie.
This paradigm shift positions platforms as consumer products, viewing developers as customers. Mature IDPs consolidate service catalogs, templates, and performance metrics such as DORA’s deployment frequency and lead time, as well as user satisfaction scores. Spotify, for instance, achieved a 55% reduction in time-to-tenth pull request after implementing Backstage. Looking ahead, PlatformEngineering.org forecasts that by 2026, AI agents will function as platform peers, with “vibe coding” necessitating platforms to act as code reviewers and automatic remediators.
As enterprises prepare for this transition, 93% are planning to expand their GitOps practices, leading platforms to evolve into AI-ready ecosystems characterized by predictive DevSecOps, Software Bill of Materials (SBOMs), and feature flags as essential components. Platform engineers are also seeing a significant pay increase, earning 27% more than their DevOps counterparts, signifying the emergence of distinct career paths alongside Site Reliability Engineering (SRE) and FinOps specialists.
By 2026, the rise of “agentic developer platforms” will enable AI agents to play a critical role throughout the software lifecycle, from intent to infrastructure and even the autonomous decommissioning of obsolete resources. “By 2026, we’ll start seeing AI agents move beyond single-company experiments and into limited cross-organization collaboration, but it will still be cautious and deliberate,” forecasted Max Marcon, MongoDB’s director of product management, in The New Stack.
However, challenges remain, particularly around agent interoperability, reminiscent of the early days of Open Banking, which necessitates the establishment of standards such as the Model Context Protocol (MCP). Without robust platforms, the integration of AI may exacerbate existing friction and risks, intensifying concerns around permissions, accountability, and infrastructure-as-code (IaC) drift detection. Platforms must therefore embed governance mechanisms, utilizing GitOps as a control plane, establishing time-to-live policies, and integrating FinOps to minimize cloud waste. Thomson Reuters, for example, achieved a 15-fold productivity increase and 70% automation with the Amazon Bedrock AgentCore for operations.
In practical applications, Spotify’s AiKA tool has improved support resolution times by 47%, while popular tools like Claude, Cursor, and GitHub Copilot are used daily by 90% of developers, as reported by DORA, all operating under carefully vetted guardrails. “AI will positively disrupt the entire software development life cycle… 10x engineers could become 100x,” predicted Paul Payne, CTO of SaaScada.
The industry is witnessing a surge in managed solutions for IDPs, such as Roadie or Port, which are expected to thrive by 2026, facilitating team focus on innovation. The market is projected to grow from $7.19 billion in 2024 to $40.17 billion by 2032. Platform Engineering Week will also feature events in Berlin from June 15–19 and Amsterdam from April 20–24, covering topics like scalable IDPs, AI governance, and return on investment measurement.
Success metrics are evolving, focusing on adoption rates, time-to-first-implementation, and data-driven scorecards. “We’re entering a phase where small teams can do previously unthinkable things,” noted Helen Greul, VP at Multiverse.io. Leadership must prioritize developer experience, as DORA links user-centric platforms to improved outcomes. “As technology evolves… AI agents become more autonomous… independently take actions,” envisioned João Freitas of PagerDuty. The fundamental principles remain: no AI should bypass testing or compliance. Organizations that balance innovation with structure—through golden paths and observability—are likely to thrive, transforming platforms into engines of AI value.
As we look towards 2026, the predictions by PlatformEngineering.org suggest a landscape where AI agents gain RBAC capabilities, vibe coding becomes a standard for remediation, and agentic platforms integrate lifecycle AI. The Cloud Native Computing Foundation (CNCF) envisions autonomous enterprises built around four pillars: intent-to-infrastructure, optimization, and governance. Discussions on the future of platform engineering highlight a shift towards backend orchestration superseding traditional portals while fostering AI-native self-healing operations.
High-performing organizations are beginning to measure both system performance and developer conditions, emphasizing the need for new roles like AI Operations and Developer Experience Engineers. GitOps is increasingly dominating pipelines, promoting auto-debugging and enhancing resilience through progressive delivery, thereby resetting traditional DevOps norms. As Meyen observes, the convergence of infrastructure, APIs, data, and AI is no longer optional; it is the essential forge defining the future of enterprise software delivery.
See also
Microsoft Launches Maia 200 AI Chips, Boasting 30% Better Performance per Dollar
Microsoft Launches Maia 200 AI Chip, Achieving 3x Inference Performance Boost
NHS England Trials AI and Robotics for Early Lung Cancer Detection, Aiming for 50,000 Diagnoses by 2035
Microsoft Launches Maia 200 Chip, Promises 30% Better AI Performance Per Dollar
