Marvelous MLOps’ cover photo
Marvelous MLOps

Marvelous MLOps

E-Learning Providers

Power up MLOps with Marvelous content

About us

Hey there! We’re Başak and Maria, friends and colleagues who share a love for Marvel and MLOps. When we decided to start a blog, we asked ChatGPT for help in finding a name, and it came up with “Marvelous MLOps” — that’s how we got our start! We’ve both been working as Machine Learning engineers in the corporate world for a while now, and we believe it’s time to share our knowledge with others. Three months later Raphaël joined us as our creative director and started onboarding by watching some Marvel movies 😊. In Marvelous MLOps, we are for pragmatic MLOps. We are for connecting existing tools in a way that supports MLOps standards and show how we do it in our blog. We’re excited to connect with fellow enthusiasts and collaborate to find solutions to any challenges we may encounter.

Website
https://www.marvelousmlops.io
Industry
E-Learning Providers
Company size
2-10 employees
Headquarters
Amsterdam
Type
Self-Employed
Founded
2023

Locations

Employees at Marvelous MLOps

Updates

  • One year of Marvelous MLOps! 🥂 A year has already passed since we officially registered Marvelous MLOps as a company. It’s been a lovely journey—launching our first course, seeing one cohort successfully complete it, and having another one in full swing. Looking ahead, year two will be even more exciting—more learning, more building, and more impact. Grateful for everyone who has been part of this journey so far! Here’s to another year of learning and collaboration 🚀

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    Building robust ML pipelines in Databricks? Here's a production-ready pattern we use to automate model retraining, evaluation, approval, and deployment 👇 1️⃣ The workflow begins with a preprocessing step that prepares the data. 2️⃣ Next, the train_model task retrains the model, evaluates its performance against the latest registered model on the new test set, and registers it if there is an improvement. It sets two task values: model_updated and model_version. 3️⃣ The model_updated task is an IF task in the Lakeflow job. It checks whether a new model was registered. If not (model_updated = 0), the job ends early. If a new model was registered (model_updated = 1), the workflow continues to approve_check 4️⃣ The approval step checks whether the tag value (the name of the tag matches the task name) is “approved”. If not, the job fails (it always fails the first time). Here, we utilize MLflow deployment jobs, currently in Beta 5️⃣ The approver must manually approve the new model version via the UI. Once done, the workflow repairs itself and moves on to the deploy_model task. 6️⃣ The deploy_model task It wraps the model as a pyfunc and deploys it to an endpoint using the recorded model_version. Each task in the workflow (except the IF task) runs a python script that requires specific parameters. Some parameters are set by the asset bundles: - “env” parameter matches the bundle target - “root_path”  is the root directory of the bundle - “git_sha” and “branch” are variables defined in the bundle that get overwritten via the CD pipeline - “run_id”, “job_id”, and “task_name” are set using dynamic value references {{job.run_id}, {{job.id}} and {{task.name}}, which get filled when the job starts running. Parameters are used to access configuration files necessary for the deployment, and tag experiments, models, and deployments so that we can always find the corresponding code and run related to them. This setup ensures that only improved and explicitly approved models are deployed to production. ------------------------------------------------------------------- 💡 Want to learn more? We teach how to build such pipelines in our Marvelous MLOps End-to-end #MLOps with #Databricks course. Link in the bio. Use code MARVELOUS for 100 euro off!

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    An LLM course that focuses on systems 👇 I will be joining the next cohort of Building LLM Applications for Data Scientists and Software Engineers by Hugo Bowne-Anderson and Stefan Krawczyk, starting on the 8th of July. It's a highly interactive course with: - hands-on projects - deep-dive into practical topics including debugging, evaluation, observability, agents, and deployment - amazing guest speakers - over $1500 in credits! I'm beyond excited! This time, lectures take place at 18:00 CET on Tuesdays and Thursdays. Join me and learn how to ship real AI systems. Use this link for 10% off: https://lnkd.in/gx8SExjs #ai #llm #softwareengineering

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    #MLOps is just #DevOps for ML... A common misconception. MLOps is much more. Traditional software systems rely on four stable pillars: ✅ Infrastructure ✅ Environment ✅ Code ✅ Data Model If these remain unchanged, performance is predictable. ❗But #AI systems break this rule. Even with no changes, models can degrade due to data distribution shifts. That’s why data versioning & continuous monitoring of data/model drift are critical for AI systems. That's what makes MLOps hard. Meme credit: ClearML

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    My current workflow developing on #Databricks 👇 - Connect to Databricks via Databricks VS code extension. - Develop in VS code with interactive cells. - Use uv for setting up virtual environment, package building (build backend setuptools, but can be hatchling) and dependency resolution. - Start writing modular code & package it as soon as possible keeping the scripts clean. - If some code can’t be executed from the VS code, sync the files (including wheel files) using VS code extension and execute in a notebook (do not forget to install the package). - When the script is ready, use asset bundle deploy & run command to execute it on interactive cluster. - Databricks.yml definition contains artifacts section that will build and upload package to the bundle location and make it accessible for the script to run. - It is important to set dynamic_version=True when running on interactive cluster. In this way, all package changes will be availble on the cluster. Here is an example of basic databricks.yml file 👇. #mlops

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Georgian Pirvu

    Cloud Lead Data & AI Engineer | 5 x Azure certified | 2 x Databricks certified Databricks Champion

    This week we just wrapped up the “End-to-End MLOps with Databricks” cohort by Maria Vechtomova and Başak Tuğçe Eskili, where at the end I had the chance to also present my MLOps solution built on Databricks! If you're serious about production-grade ML on Databricks, I highly recommend this course: 👉 https://lnkd.in/e3Aj2fHg 💡 The 7-week journey is packed with insights on how to implement MLOps the right way — focusing not just on tools, but on applying engineering and operational principles that make models production-ready, traceable, and maintainable. From Databricks Asset Bundles, Unity Catalog, feature serving, to CI/CD integration, the course covered how to turn fragmented ML experimentation into a robust, collaborative, and automated workflow. 💡 What you will appreciate most: Emphasis on reproducibility: link every model to code, data, infra, and environment. Community-based feedback and hands-on project showcase Direct feedback loop with the cohort owner. 📌 Huge thanks to Maria and Başak for building one of the most practical MLOps courses out there and the journey will continue with AI Engineering. Feel free to reach out to them. #MLOps #Databricks #MachineLearning #DataEngineering

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    "AI is not a bubble". Says everyone who is deeply invested in keeping the momentum alive. Bubbles in technology are not new. In 2007, an a capella group The Richter Scales created a song "Here comes another bubble". They received an Webby Award for Viral Video. Today, this video is more relevant than ever! I decided to create a parody on this song. Lyrics is mostly written by me (ChatGPT is not great in understanding the beat). The song is made with AI and edited by me. The world must see it. Like, comment and reshare! Follow Marvelous MLOps on Youtube not to miss our free upcoming course on MLOps with Databricks (launching on the 28th of July). 🔗 https://lnkd.in/eszY9qEx

    Here comes another bubble (2025)

    https://www.youtube.com/

  • View organization page for Marvelous MLOps

    14,978 followers

    Almost got a CS degree Stanford dropout, dreaming free Watched the hype and took the bait Now I’m coding #AI fate November thirtieth, it dropped ChatGPT — the world has stopped Everyone just paused and stared Google watched — but wasn't scared GPT-4 made its mark Claude appeared, polite and smart Google scrambled, shipped out Bard (Launching LLMs is hard) Left my crypto days behind Bragged I’ve built a thinking mind AI agent, it's so hot With MCP — I kid you not Dreamed I'd close a massive seed "AGI" was all I'd need Even though it’s held with tape VCs came to seal the fate Here comes another bubble The VCs are buzzing AI does everything (but nothing) Launched a demo late one night Woke up, X has lost its mind Angels circled, term sheets flew “Pre-revenue? We love that too” Did demo live, it didn’t crash Woke up trending, raised more cash Forbes 30 Under 30 call Didn’t know I’d peaked at all Blog it, film it, make those reels TikTok trends and viral feels Tweet it, thread it, demo day Build a bot that writes your way Here comes another bubble In a year, we’ll pivot AGI — we live it “Helps you code” but wrecks your build Kills the flow, your patience spilled System broken, things unclear Keep believing, AGI is near Billion tokens burned in vain Models hallucinate again Now your startup’s pivoted To crème brûlée with AI flame Here comes another bubble And when we are gone This will still go on And on and on and on And on and on POP!

  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    It is an honor to be featured among top 100 courses on Maven! Over 150 machine learning engineers, data scientists and architects have already built end-to end #machinelearning projects with Marvelous MLOps! What is special about the course (and we are proud of it!): - By the end of the course, you will know how to apply #MLOps principles for real-life projects - You will get lifetime access to the course materials and updates. - We review your PRs and are always available on Discord (also after the course!) - This is the only practical course on MLOps I know about that goes this deep! - You get access to a Databricks environment (included in the price). This is the last time Başak and I will be teaching this specific course. This week, Maven provides a 20% discount for the top 100 courses on the platform. Do not miss this unique opportunity: https://lnkd.in/e34GVVnk

    • No alternative text description for this image
  • Marvelous MLOps reposted this

    View profile for Maria Vechtomova

    MLOps Tech Lead | Writing a book for O'Reilly | Databricks MVP

    Top FREE or affordable resources to learn about #MLOps and #AI 𝗕𝗼𝗼𝗸𝘀: 📕Designing Machine Learning Systems 𝖻𝗒 Chip Huyen 📕AI Engineering: Building Applications with Foundation Models by Chip Huyen 📕LLM Engineer's Handbook by Maxime Labonne and Paul Iusztin 📕Building LLMs for Production by Louis-François Bouchard 📕Machine Learning Engineering with Python by Andy McMahon 𝗡𝗲𝘄𝘀𝗹𝗲𝘁𝘁𝗲𝗿𝘀: 📝 Decoding ML 𝖻𝗒 Paul Iusztin: https://lnkd.in/ee-cd7PB 📝The Neural Maze by Miguel Otero Pedrido: https://lnkd.in/eHNMJ3W8 📝 Real-World Machine Learning by Pau Labarta Bajo: https://lnkd.in/eNcJmMW6 📝 Marvelous MLOps Substack: https://lnkd.in/eZ8GbF5x 𝗣𝗼𝗱𝗰𝗮𝘀𝘁𝘀: 🎙Vanishing Gradients by Hugo Bowne-Anderson: https://lnkd.in/eP8TSqnX 🎙MLOps community podcast by Demetrios Brinkmann: https://lnkd.in/eD-9ZERv What are your favourite learning materials?

    • No alternative text description for this image

Similar pages

Browse jobs