CNCF와 SlashData 보고서, 플랫폼 엔지니어링 전략이 AI 인프라 도입을 형성하고 있다고 밝혀

Latest research finds 54% of developers with independently managed tooling use hybrid approaches for AI workflows, compared to just 5% at organizations building custom internal platforms

Key Highlights:

CNCF and SlashData released the Q3 2026 CNCF Technology Radar report during KubeCon + CloudNativeCon North America 2026, examining developer perceptions of observability, infrastructure as code and storage technologies.

Prometheus and OpenTelemetry lead observability, earning 'Adopt' positions based on strong developer confidence in their maturity, usefulness and reliability.

OpenTofu and OpenFeature lead infrastructure as code, while Carina earns 'Adopt' positions in storage, highlighting developer confidence across cloud native infrastructure.

Platform engineering ownership is shaping AI infrastructure strategies, with 54% of developers at organizations with independently managed tooling reporting hybrid AI workflows, compared to just 5% at organizations building custom internal developer platforms.

, /PRNewswire/ -- The Cloud Native Computing Foundation® (CNCF®), which builds sustainable ecosystems for cloud native software, today released new findings from its Q3 2026 CNCF Technology Radar report , developed in collaboration with SlashData . As organizations expand their use of artificial intelligence (AI), the report highlights how platform engineering strategies influence AI infrastructure decisions and identifies the cloud native technologies developers view as mature, useful and reliable.

The research examines developer perceptions of observability, infrastructure as code (IaC) and storage technologies, identifying tools considered reliable choices for broad adoption. Findings also reveal how centralized and distributed platform engineering models influence whether organizations extend existing platforms, build dedicated AI infrastructure or adopt hybrid approaches to managing AI workflows.

"Cloud native technologies are foundational to building and operating modern infrastructure, and AI is introducing new demands around scalability and reliability," said Chris Aniszczyk, CTO, CNCF. "What this research reinforces is that platform engineering is playing a critical role in how organizations approach AI infrastructure. As teams move AI workloads into production, mature CNCF projects like Backstage, Prometheus and OpenTelemetry provide the visibility and reliability needed to support complex environments."

Platform engineering ownership shapes AI infrastructure strategies

As organizations integrate AI workflows into their infrastructure, the report finds that platform ownership is closely associated with how teams approach AI infrastructure. Organizations with centralized platform engineering teams tend to commit to a single strategy, while those with distributed ownership are more likely to adopt hybrid approaches.

Key findings include:

54% of developers at organizations where individual teams independently manage their tooling use hybrid AI approaches , compared to 43% at organizations where multiple teams collaboratively provide platform capabilities.

38% of developers at organizations building custom internal developer platforms report using a separate, dedicated AI platform , while 33% extend their existing developer platform to support AI workflows.

Only 5% of developers at organizations building custom internal platforms use a hybrid approach , compared to 18% at organizations with centralized teams curating third-party tools.

25% of developers at organizations with platform teams curating third-party tools rely primarily on AI vendor platforms , the highest proportion among the platform ownership models surveyed.

These findings suggest that centralized platform teams are better positioned to establish consistent infrastructure strategies, while distributed ownership can lead to more fragmented approaches to AI deployment and management.

As AI workloads move toward production, the ability to standardize infrastructure and coordinate platform capabilities will become important for managing operational complexity, resource efficiency and reliability.

"The data shows that how organizations structure their platform engineering teams has a meaningful relationship with how they approach AI infrastructure," said Liam Bollmann-Dodd, principal market research consultant at SlashData. "Centralized teams tend to make deliberate decisions about extending existing platforms or building dedicated AI environments, while distributed teams are more likely to adopt hybrid approaches. This highlights how organizational structure can influence the complexity of supporting AI workloads in production."

Prometheus and OpenTelemetry lead developer confidence in observability

As cloud native environments grow more complex, observability remains an important component of reliable infrastructure. The report finds that Prometheus and OpenTelemetry lead the observability category, earning 'Adopt' positions based on developer perceptions of maturity, usefulness and likelihood to recommend.

The report also highlights strong developer recommendations for Fluent Bit, Jaeger and Cortex, demonstrating confidence in tools supporting monitoring, logging and distributed tracing.

Key findings include:

Prometheus received four- or five-star maturity ratings from 80% of developers surveyed , including 38% who awarded it five stars for stability and reliability.

OpenTelemetry received five-star usefulness ratings from 34% of developers , the second-highest proportion among observability technologies. Additionally, 92% of current and former users were likely or highly likely to recommend it.

99% of current and former Fluent Bit users were likely or highly likely to recommend it , making it the most recommended observability technology surveyed, despite more moderate maturity and usefulness ratings.