We're sunsetting PodQuest on 2025-07-28. Thank you for your support!
Export Podcast Subscriptions
cover of episode MLOps with Databricks // Maria Vechtomova // #314

MLOps with Databricks // Maria Vechtomova // #314

2025/5/13
logo of podcast MLOps.community

MLOps.community

AI Deep Dive Transcript
People
D
Demetrios
M
Maria
Topics
Maria: 我在MLOps领域拥有超过八年的经验,亲身经历了各种工具和平台的演变。早期,我们不得不自建工具来完成模型注册、实验跟踪等任务。现在,虽然市面上有很多优秀的工具,但将它们整合在一起非常复杂,需要投入大量资源和人力。因此,对于大型云端组织来说,选择一个集成的平台可能更为高效。Databricks正是一个备受推崇的平台,它提供了MLOps所需的各种组件,虽然并非完美,但足以满足大多数需求。我尤其欣赏Databricks对用户反馈的开放态度,以及他们不断改进的决心。然而,Databricks的开发流程仍然存在一些痛点,例如过度依赖notebooks,以及Feature Store的一些设计缺陷。我极力推荐使用Asset Bundle进行开发,它可以更好地管理代码和依赖项。总的来说,Databricks是一个功能强大且不断进步的平台,但用户需要了解其优缺点,并根据自身情况做出明智的选择。 Demetrios: 我与Maria的讨论深入探讨了Databricks在MLOps中的应用,涵盖了其优点、缺点和不足。我们一致认为,对于许多组织来说,选择一个集成的平台而非拼凑各种工具,可以大大简化MLOps流程。Databricks的普及程度和不断增长的能力使其成为一个有吸引力的选择,尤其是在数据工程已经在使用Databricks的情况下。然而,我们也指出了Databricks的一些局限性,例如开发流程的痛点和Feature Store的设计缺陷。Maria分享了她在Databricks上的实践经验和最佳实践,为其他用户提供了宝贵的参考。总的来说,Databricks是一个强大的平台,但用户需要了解其优缺点,并根据自身情况做出明智的选择。

Deep Dive

Shownotes Transcript

MLOps with Databricks // MLOps Podcast #314 with Maria Vechtomova, MLOps Tech Lead | Founder at Ahold Delhaize | Marvelous MLOps.

Join the Community: https://go.mlops.community/YTJoinIn

Get the newsletter: https://go.mlops.community/YTNewsletter

// Abstract

The world of MLOps is very complex as there is an endless amount of tools serving its purpose, and it is very hard to get your head around it. Instead of combining various tools and managing them, it may make sense to opt for a platform instead. Databricks is a leading platform for MLOps. In this discussion, I will explain why it is the case, and walk you through Databricks MLOps features.

// Bio

Maria is an MLOps Tech lead with over 10 years of experience in Data and AI.

For the last 8 years, Maria has focused on MLOps and helped to establish MLOps best practices at large corporations.

Together with her colleague, she co-founded Marvelous MLOps to share knowledge on MLOps via training, social media posts, and blogs.

// Related Links

Website: marvelousmlops.io

MLOps Course discount code: MLOPS100 for the podcast listeners - https://maven.com/marvelousmlops/mlops-with-databricks?promoCode=MLOPS100


Catch all episodes, blogs, newsletters, and more: https://go.mlops.community/TYExplore

Join our slack community [https://go.mlops.community/slack]

Follow us on X/Twitter [@mlopscommunity](https://x.com/mlopscommunity) or [LinkedIn](https://go.mlops.community/linkedin)]

 Sign up for the next meetup: [https://go.mlops.community/register]

MLOps Swag/Merch: [https://shop.mlops.community/]

Connect with Demetrios on LinkedIn: /dpbrinkm

Connect with Maria on LinkedIn: /maria-vechtomovaTimestamps:

[00:00] Maria's preferred coffee[00:42] Takeaways[02:48] Why Databricks for MLOps[09:56] Platform Adoption vs Procurement Pain[12:56] Databricks Best Practices[16:57] Feature Store Overview[22:00] Managed system trade-offs[29:15] Databricks Developments and Trends[44:31] Insider Info and Summit[45:47] Data Ownership Pros and Cons[48:08] Data Contracts and Challenges[51:25] MLOps Databricks Book Guide[52:19] Wrap up