# LUML > Open-source MLOps and LLMOps platform. Registry, deployment, monitoring and agent observability, on infrastructure you own. ## Pages - [LUML](https://luml.ai/): Open-source MLOps and LLMOps platform. Registry, deployment, monitoring and agent observability, on infrastructure you own. - [LUML Core](https://luml.ai/core): Registry, deployment and monitoring inside your own infrastructure. - [LUML Flow](https://luml.ai/flow): Local experiment tracking for ML runs, GenAI traces and evals. - [LUML Prisma](https://luml.ai/prisma): Autonomous research module that optimises ML pipelines. - [LUML pricing](https://luml.ai/pricing): Plans that scale with your team. Every tier runs the same open-source platform. ## Per-page detail - https://luml.ai/core/llms.txt - https://luml.ai/flow/llms.txt - https://luml.ai/prisma/llms.txt - https://luml.ai/pricing/llms.txt - https://luml.ai/blog/llms.txt --- I work on Predictive ML Generative AI ## The open-source MLOps platform ### Where engineers and agents work together. You set the objective. Agents run the experiments. Every run is tracked and reproducible, one click from production, and monitored once it's there. [Get started (free)](https://app.luml.ai)[Request demo](https://calendar.app.google/pb2b93ULjATupqVS7) ## The Problem: ### Everyone wants agents doing ML work. Nobody has a clear structure. The capability isn't the gap. Agents already write the training code, run the experiments, and can carry a model the whole way to production. What's missing is the structure around that — where runs get tracked, how they get compared, explained, reproduced, and productionized. So every team improvises: different ways of working, different tools, wired to agents differently each time. It holds up until you have to explain a result, rerun it, or hand it to someone else. ## The Solution: ### We have a structure. It is called LUML . Five steps, from your first instruction to a model running in production. STEP 01 / 05 ### Work with agents, side by side You drive a coding agent from the terminal you already use. What it produces appears beside you as work you can look at: experiments, metrics, models. Whether you trained it or the agent did, it is captured the same way. STEP 02 / 05 ### The agents get better at your domain None of that session is thrown away. What worked, what you overruled, and the details specific to your domain get consolidated into your team's context — so the next session starts from the knowledge you already provided, instead of from scratch. STEP 03 / 05 ### Hand over the objective Once the context holds enough, you can stop supervising each run and state a goal instead. Prisma forks it into parallel branches, runs each in its own git worktree and proposes the next direction. Weak branches are pruned, a fixed metric defines the winner, and the winning model comes back to you for review. STEP 04 / 05 ### The winning model folds into one artifact Weights, metadata, dependencies and preprocessing fold into a single self-contained .luml artifact. The registry versions it and links it back to the exact experiment that produced it. Nothing about how it was made is left behind on a laptop. STEP 05 / 05 ### Deployed and monitored A service you host inside your own network reaches out, pulls the artifact and serves it, so nothing has to reach in. Monitoring sits on the same machine, watching drift, data quality and health where the model actually runs. ## Meet the Team ### We built the tool we always needed We're data scientists and engineers who've built AI systems end-to-end at L'Oréal · Bitpanda · SCCH — across startups and large-scale environments. Iryna CEO Oleh CTO Kate Backend Developer Olha Data Scientist Valera Frontend Developer Anna UI/UX Designer Nikita Technical Advisor ## FAQ ### Frequently asked questions What does LUML cover end-to-end? The path from first experiment to production, for humans and agents on one platform: experiment tracking with traces and evals, autonomous research loops, and a registry with deployments and monitoring on your own infrastructure. Do agents replace my team? No. Agents execute and search; your team sets objectives, reviews outcomes, and keeps the decisions. LUML's job is making agent work reviewable: versioned, comparable, and deployable instead of buried in chat history. Which coding agents does LUML work with? Prisma drives the popular coding CLIs: Claude Code, Codex, Gemini CLI, Cursor, Copilot CLI, and opencode. It orchestrates them locally and calls no LLM APIs itself. Is LUML open-source? Yes. The core platform is open-source under a permissive license. You can inspect the code, contribute, and self-host if needed. Where does my data live? In your cloud storage. LUML connects to your S3 or Azure buckets but never stores your artifacts or data. Everything stays where you put it. How is LUML different from MLflow / Weights & Biases? Flow is a local-first tracker with an MLflow drop-in plugin: point existing code at luml://local and it keeps working, with LLM traces and evals included. From there the same platform adds autonomous experimentation and the path to production, which trackers alone don't cover. How is LUML different from SageMaker / Vertex AI / Azure ML? Cloud ML platforms lock you into a single provider and come with significant complexity. LUML is cloud-agnostic: connect your own storage on AWS, Azure, or anywhere else. It’s simpler to set up, easier to use day-to-day, and you keep full control over your infrastructure. How is LUML different from Langfuse / LangSmith / Arize? Those tools focus primarily on LLM observability. LUML covers both ML and LLM workloads across the full lifecycle, from experiment tracking through deployment, in a single platform. Can we adopt gradually? Yes. Start with Flow locally on one project, add the registry and deployments when you want to ship, and bring in Prisma when a problem deserves a search. No migration required. How quickly can we get started? Minutes. Run pip install lumlflow to track locally with no account, or create a free account, connect your cloud storage, and register your first model with the SDK. What about pricing? Light is free forever and includes all platform capabilities for one user, one orbit and 50 artifacts. Paid plans are priced per workspace rather than per seat: Plus from €79 a month, Pro from €249 and Scale from €599, each on annual billing, or €99, €299 and €699 month to month. Enterprise is a custom annual contract. See the pricing page for what each tier includes. ### One platform for the entire AI lifecycle. Open-source and free to start. [Get started (free)](https://app.luml.ai)[Request demo](https://calendar.app.google/pb2b93ULjATupqVS7)