--- title: How we run Tuist with AI agents, Kubernetes, and Elixir description: A company describes how it rebuilt its infrastructure and operations around interfaces that AI agents can directly inspect and manipulate, moving from a PaaS... canonical: https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: How we run Tuist with AI agents, Kubernetes, and Elixir | daily.dev og:description: A company describes how it rebuilt its infrastructure and operations around interfaces that AI agents can directly inspect and manipulate, moving from a PaaS... og:url: https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4 og:image: https://api.daily.dev/og/posts/Ak6MEydt4.png og:image:alt: How we run Tuist with AI agents, Kubernetes, and Elixir og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # How we run Tuist with AI agents, Kubernetes, and Elixir **[ Tuist ](https://daily.dev/sources/tuist-blog)** · 14 min read · 1 upvotes · 0 comments ## Summary A company describes how it rebuilt its infrastructure and operations around interfaces that AI agents can directly inspect and manipulate, moving from a PaaS to Kubernetes for flexible, agent-debuggable infrastructure, leveraging Elixir/Erlang's runtime introspection for production debugging, using Grafana with Prometheus/OpenTelemetry in a largely headless (agent-queried rather than human-dashboard-driven) way, and extending the same philosophy to company operations via an internal context system called Atlas accessed through MCP. The overarching argument is that adopting established, standardized 'languages' (Kubernetes, Erlang, OpenTelemetry) rather than convenient but opaque abstractions lets agents understand and act on systems at every level of the organization. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### Why would a team move from a PaaS like Fly.io to Kubernetes specifically to work better with AI agents? Kubernetes is an established, well-documented language that large language models already understand deeply, so agents can inspect and modify infrastructure state without needing an interpreting layer like MCP. This gives teams low-level control over latency and resource usage, the ability to port across cloud providers, and lets agents debug infrastructure state directly through Kubernetes' runtime interfaces instead of a PaaS UI. _daily.dev surfaces these kinds of real-world Kubernetes and agent-ops tradeoffs for teams rethinking their infrastructure stack._ ### How can AI agents help debug a misbehaving Elixir application in production? Agents can open a console directly against the running Erlang VM, inspecting all live processes and their symbols, sorting them by memory or CPU usage, and locating specific processes by identifier. Because the Erlang scheduler is fair, this introspection works even under heavy production traffic, letting an agent diagnose issues without needing a separate reproduction environment. _Engineers debugging Elixir or Erlang systems can find more operational deep dives like this through daily.dev._ ### What is Pulso and how does it relate to Grafana for observability? Pulso is a stateless, object-storage-backed tool built in Elixir with Rust for performance-critical parts, designed for storing and querying observability data and sending alerts. It is being developed as a work-in-progress replacement for hosted Grafana, aiming to give the team full operational control, scalability, and the ability to improve the system themselves rather than depending on a third-party observability vendor. _daily.dev helps teams evaluating observability stacks like Grafana versus self-built alternatives stay on top of new tools._ ## Similar posts on daily.dev - [Kubernetes Was the Easy Part](https://daily.dev/posts/kubernetes-was-the-easy-part-kzq9zt3ac) · Cloud Native Now · 1 upvotes · 0 comments - [Building an Internal Developer Platform with Artificial Intelligence](https://daily.dev/posts/building-an-internal-developer-platform-with-artificial-intelligence-arvvvck6u) · InfoQ · 1 upvotes · 0 comments - [Automating Infrastructure as Code changes with an AI agent](https://daily.dev/posts/automating-infrastructure-as-code-changes-with-an-ai-agent-orguqzhbz) · CircleCI · 0 upvotes · 0 comments - [IEEE Cloud Summit 2026: The Tunnels No One Mapped](https://daily.dev/posts/ieee-cloud-summit-2026-the-tunnels-no-one-mapped-livjy4s3b) · GitGuardian · 1 upvotes · 0 comments --- Tags: [#ai-agents](https://daily.dev/tags/ai-agents), [#kubernetes](https://daily.dev/tags/kubernetes), [#observability](https://daily.dev/tags/observability), [#elixir](https://daily.dev/tags/elixir), [#grafana](https://daily.dev/tags/grafana) [View this post on daily.dev](https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"How we run Tuist with AI agents, Kubernetes, and Elixir","url":"https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4"},"datePublished":"2026-10-02T19:51:26.989Z","dateModified":"2026-10-02T19:51:50.606Z","description":"A company describes how it rebuilt its infrastructure and operations around interfaces that AI agents can directly inspect and manipulate, moving from a PaaS...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/61fa748c6e8083e34f91e8dcc5bd35c9?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/61fa748c6e8083e34f91e8dcc5bd35c9?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"\n Tuist\n ","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"\n Tuist\n ","logo":"https://media.daily.dev/image/upload/s--3AYDH23h--/f_auto,q_auto/v1787487506/logos/tuist-blog?_a=BAMAMicg0","url":"https://daily.dev/sources/tuist-blog"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai-agents,kubernetes,observability,elixir,grafana","timeRequired":"PT14M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"\n Tuist\n ","item":"https://daily.dev/sources/tuist-blog"},{"@type":"ListItem","position":3,"name":"How we run Tuist with AI agents, Kubernetes, and Elixir"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/how-we-run-tuist-with-ai-agents-kubernetes-and-elixir-ak6meydt4#faq","mainEntity":[{"@type":"Question","name":"Why would a team move from a PaaS like Fly.io to Kubernetes specifically to work better with AI agents?","acceptedAnswer":{"@type":"Answer","text":"Kubernetes is an established, well-documented language that large language models already understand deeply, so agents can inspect and modify infrastructure state without needing an interpreting layer like MCP. This gives teams low-level control over latency and resource usage, the ability to port across cloud providers, and lets agents debug infrastructure state directly through Kubernetes' runtime interfaces instead of a PaaS UI. daily.dev surfaces these kinds of real-world Kubernetes and agent-ops tradeoffs for teams rethinking their infrastructure stack."}},{"@type":"Question","name":"How can AI agents help debug a misbehaving Elixir application in production?","acceptedAnswer":{"@type":"Answer","text":"Agents can open a console directly against the running Erlang VM, inspecting all live processes and their symbols, sorting them by memory or CPU usage, and locating specific processes by identifier. Because the Erlang scheduler is fair, this introspection works even under heavy production traffic, letting an agent diagnose issues without needing a separate reproduction environment. Engineers debugging Elixir or Erlang systems can find more operational deep dives like this through daily.dev."}},{"@type":"Question","name":"What is Pulso and how does it relate to Grafana for observability?","acceptedAnswer":{"@type":"Answer","text":"Pulso is a stateless, object-storage-backed tool built in Elixir with Rust for performance-critical parts, designed for storing and querying observability data and sending alerts. It is being developed as a work-in-progress replacement for hosted Grafana, aiming to give the team full operational control, scalability, and the ability to improve the system themselves rather than depending on a third-party observability vendor. daily.dev helps teams evaluating observability stacks like Grafana versus self-built alternatives stay on top of new tools."}}]} ```