AS
Anshul Sao
@anshulsaofacets
ClickOps with LLMs Won't Scale: Why You Need AI-Powered Platform-Led Orchestration
Submitted Apr 16, 2025
Type of submission:
30 mins talk
Topic of your submission:
AI infrastructure and operations
I am submitting for:
Rootconf Annual Conference 2025
Abstract
LLMs generating Terraform sounds magical until pipelines fail, environments drift, and your infra becomes unmanageable. This talk critiques the limitations of AI-based ClickOps and proposes a better approach: AI-powered platform-led orchestration. Learn how to structure internal platforms using APIs, reusable modules, and environment-aware workflows that deliver velocity without sacrificing governance.
ClickOps is back with GenAI tools that turn prompts into infrastructure. But under the hood, these tools often generate fragile, unstructured scripts. Terraform state grows uncontrollably, pipelines break on unrelated changes, and governance is reactive at best.
This talk challenges the assumption that LLMs alone can solve infrastructure delivery. While AI can assist with discovery and mapping, real control comes from a platform-led model: one where infrastructure is consumed via structured APIs, built from versioned modules, and deployed using environment-aware workflows.
Takeaways
-
The pitfalls of LLM-generated infrastructure: drift, duplication, and deployment fragility
-
Why structured APIs, versioned modules, and declarative workflows are non-negotiable
-
How to use AI within a platform not as a replacement for it to improve discovery, blueprinting, and governance
Who Can Benefit From Talk/Submission?
This session is ideal for Platform Engineering and DevOps leaders looking to scale infrastructure delivery without compromising control, as well as CTOs, Cloud Architects, and SREs seeking practical strategies to integrate AI tooling within governed, scalable platforms. If you’re building internal platforms or managing environment consistency across clouds, this talk is for you.
Speaker’s Bio
Anshul Sao is the CTO and Co-founder of Facets.cloud, where he helps organizations transition from DevOps chaos to platform-led infrastructure orchestration. With a deep background in infrastructure automation, Anshul’s recent work focuses on integrating AI and LLMs into internal developer platforms to streamline environment provisioning and reduce operational friction. Prior to Facets, he served as Chief Architect at Capillary Technologies, leading cloud modernization initiatives and building internal tooling for multi-region deployments using Kubernetes, Terraform, and Helm.
{{ gettext('Login to leave a comment') }}
{{ gettext('Post a comment…') }}{{ errorMsg }}
{{ gettext('No comments posted yet') }}