OpsMill’s cover photo
OpsMill

OpsMill

Software Development

Great infrastructure automation starts with great infrastructure data.

About us

Infrahub from OpsMill is the data management platform for powering reliable infrastructure automation at scale. Just like a car won’t get far with a faulty engine, your automation won’t succeed without reliable data. When infrastructure data is fragmented or out of date, automation becomes hard to scale and harder to believe in. Infrahub brings structure and versioning to your infrastructure data, so you can build automation that’s roadworthy from the start and resilient over time. UNIFY INFRASTRUCTURE DATA • Keep all data sources in sync across teams and systems • Model your infrastructure your way with full schema control • Connect business logic to technical data to capture and build on design intent • Audit every change, who made it, and how it affects the rest of your infrastructure VERSION & VALIDATE EVERYTHING • Track every change with Git-like version control • Work safely in branches and test before you deploy • Collaborate through peer review to verify changes • Validate updates automatically with native CI workflows PACKAGE & DEPLOY AS YOU WANT • Generate configs and artifacts in a scalable way • Push artifacts to any deployment or orchestration tool in your stack • Turn infrastructure designs into versioned code that can evolve over time • Expose infrastructure as APIs to deliver X-as-a-service

Website
https://www.opsmill.com
Industry
Software Development
Company size
11-50 employees
Headquarters
Paris
Type
Privately Held
Founded
2023
Specialties
infrastructure automation, network automation, and infrastructure data management

Locations

Employees at OpsMill

Updates

  • OpsMill reposted this

    [EVENT] Pour une démarche d'automatisation plus efficace, auditable et scalable : changeons nos pratiques.   Tout part de l'intention.    Pour mieux l'appliquer, il faut d'abord définir correctement le terme de "sources de vérité".   Pour comprendre pourquoi (et comment) il est crucial de modéliser l'intention (un format clair, des consignes documentées, des configurations interprétables) :    Participez à l'édition 2 des Rencontres.   La semaine dernière, nous dévoilions les premiers speakers de cette édition 👉 Aurélien Marty (Airbus), Valentin R. (CNES) et Christophe Herail (Orange Business).   Aujourd'hui, place aux derniers speakers qui partageront leurs expériences pour interroger ce terme et déployer des approches d'automatisation efficaces, où l'intention documentée est au cœur de l'automatisation.   👉 Arnaud LE GALL (HPE Aruba Networking), Samir Medjiah (Université de Toulouse), Damien Garros (OpsMill) et Paul Tournu (CNS Communications)   Assistez à leurs talks et échangez avec eux en direct :   🟦 Préinscrivez-vous (événement gratuit) https://lnkd.in/eGEHRcz2 📅 Le 27 janvier 2026 (13h00 - 19h30) à Toulouse 🔎 Au programme : accueil déjeunatoire, interventions REX, échanges libres, soirée networking. --- LeNetDevOps, une initiative indépendante proposée par CNS Communications pour rassembler l’automatisation réseau francophone. Rejoignez la communauté sur le Slack ouvert à tous [https://lenetdevops.fr]

  • Automating a NVIDIA-based AI data center from the ground up: the true story is coming next week. Solutions Architects Alex Gittings and Mikhail Yohman will walk through how they delivered quickly on an infrastructure build while keeping flexibility and avoiding vendor lock-in. Yes, you'll get the recording afterwards if you register. 😉 But only those on the live call also get the chance to participate in the open Q+A. Bring your questions and pick our brains! 🧠 Thu, Jan 22 9am PT / Noon ET / 6pm CET Register here: https://lnkd.in/gXFYX48E

    • No alternative text description for this image
  • New year, new Infrahub! We recently released Infrahub 1.6 and 1.7, bringing improvements in usability, data governance, and platform reliability to help infrastructure teams operate at scale with less friction. A few highlights: 🔎 Audit metadata on every object (1.7): Metadata now tracks who created or updated each object and when, providing accountability and making reviews, debugging, and governance workflows easier. 🔗 Profiles can now support relationships in addition to attributes (1.7): This helps you express complete infrastructure patterns in one place and makes it easier to apply the same pattern consistently across many objects. 🎛️ Interactive dashboard (1.6): A new home page surfaces proposed changes, Git sync status, branches, recent activity, and tasks in one view, making it easier to triage reviews. 🔀 Branch management improvements (1.6): A new branch report provides visibility to ease maintenance decisions, while support for selective synchronization of Git branches and explicit merge commits gives teams more control over Git-backed workflows. In 1.7, we also upgraded critical backend dependencies to ensure optimal performance, security, and long-term maintainability. More details on the blog: 🗒️ Infrahub 1.6: https://lnkd.in/gEBxPum5 🗒️ Infrahub 1.7: https://lnkd.in/gxcfgsMm

    • No alternative text description for this image
  • Event-driven automation might sound complex but it's simply the execution of a pattern we know well: > If this, then that > If A happens, do B "That" and "B" happen without humans in the loop. So how do you build an EDA workflow into your system? The latest tutorial from Suresh Vina gives you a step-by-step on setting it up. We've used Infrahub and Ansible in the example to handle an interface status change. When the status changes in Infrahub, Ansible automatically updates the device configuration. We kept the example simple on purpose, to get you started, but the pattern can be extended to any automation task where a change in one system should trigger an action in another. The full walkthrough includes working code and a GitHub repo to build this yourself. Try it: https://lnkd.in/g4ZRH6YT

    • No alternative text description for this image
  • Look, we know what you're thinking. Another newsletter? Hard pass. And we get it. If we had a nickel for every vendor newsletter promising "game-changing insights" that was really just thinly veiled product spam... well, we'd have a lot of nickels. It's why we built the Ottermatic newsletter differently. Every other Tuesday, our team rounds up what actually matters in infrastructure automation. The stuff we're using or talking about. Battle-tested strategies. Tools worth your time. Real conversations with people who are deep in the trenches. We filter out the noise to give you the signal. One reader called it "a rare blend of practical tips and deeper thoughts." That's exactly what we're going for! If you're working in infrastructure automation—whether you're hands-on keyboard or making the business case for it—you might actually like this one. Worth a shot? https://lnkd.in/giumZ-an

    • No alternative text description for this image
  • OpsMill reposted this

    A schema-driven approach to infrastructure data helps you move fast and not break things. Especially with agentic operations. To move with speed at scale, while preserving reliability and trust. That's the point of infrastructure automation. Speed is so important, but without good control, that leads to risk, drift, and even outages. Not long ago, Damien Garros from OpsMill wrote a great blog about how a schema-driven approach to infrastructure data enables AI agents. And one of the points he makes is that the rise of AI agents brings about a new urgency to the structure and predictability of data. The reason is that agents don't just observe. They take actions. They can reason. They can plan. They can ultimately make changes in the infrastructure But unlike humans, AI agents can't "figure it out" from messy data or tribal knowledge. They need a certain degree of structure, clarity, and particularly context. They need to be able to predictably understand what the data is supposed to mean. So his point is that schema-driven infrastructure data gives AI agents a map, as opposed a maze that they'll possibly get lost in The schema defines how things work, what can change, what's safe to touch. It provides a lot of metadata context so agents can understand what the data is telling them. A schema-driven approach also allows proper scoping so that the agents know what amount of data to look at versus having a just big old blob of data that they have to search through. The point is that when infrastructure is modeled this way in a versioned, auditable, and predictable manner, AI becomes a very safe way to force multiply. And that's how you can, with AI agent enablement, move fast and not break things.

  • On Jan 22, we’re sharing real-world lessons from automating a complete NVIDIA AI data center from the ground up. From first architectural decisions to first deliverables, Solutions Architects Alex Gittings and Mikhail Yohman will walk through how to deliver quickly on an infrastructure build while keeping flexibility and avoiding vendor lock-in. Expect practical tactics you can apply now to your own automation projects. We’ll also share a real-world timeline of key milestones. Save your seat for this live session with open Q+A. Register here: https://lnkd.in/gXFYX48E Be there or be tubular. 😉

    • No alternative text description for this image

Similar pages

Browse jobs

Funding

OpsMill 1 total round

Last Round

Seed
See more info on crunchbase