Platform Engineering Rules, now with AI
Key Takeaways
The video discusses the convergence of platform engineering and AI, highlighting the importance of AI platforms and their impact on infrastructure and data communities, with tools such as Kubernetes, Prometheus, and Envoy being used to support generative AI workloads.
Full Transcript
[Music] you're watching the new sack makers a podcast for people who develop deploy and manage at scale software for more information and articles about at scale Technologies please visit the new stack. now enjoy the show the cloud native Computing Foundation or cncf hosts critical components of the global technology infrastructure including kubernetes Prometheus and Envoy cncf serves as the neutral for collaboration bringing together the industry's top developers end users and vendors and running the largest open source developer conferences hey everyone what will be the big news at cubec con this year that's always our question that the news that you what's going to be hot what are the trends that are new what are the trends that are maturing so I'm here today with priyanca Sharma executive director of the cloud native Computing foundation and Joseph sandal Who is the principal product manager platform engineering at Adobe and we're going to discuss the cloud native world and what is really hot and I was chatting a little bit beforehand with prianka and I was thinking about cubec Con in Paris which earlier I said was last year but it was actually this year and I really came away with new insights about how you're how and Engineers will have to bridge this world of data that's integrating more deeply into the infrastructure software itself and we saw new emergent uh projects such as Ray and you know and other ways we think about data and also the the the different personas that we're seeing new to the community and so I wanted to ask pranka and Joseph for a little retrospective on what they took away from those shows because that I think provides good context for what we're coming up to now absolutely hello everyone so nice to be here thank you for having us Alex um Paris yes it does feel like a lifetime ago but it was in I think yeah end of March so barely six months um I thought that was a show with unprecedented energy it was the largest cubec con ever and it really started it's it really bridged the Gap that in the past has existed between infrastructure communities data communities and AI communities that is by no means to say that the work's done let's go home but rather the work's started now so uh I found it was so much fun talking to folks like mistol folks from Gemma folks from AMA on the keynote stage these are all names in the AI space I was talking to all these people on stage and all of them without exception are dealing with infrastructure challenges and so the writing is on the wall that when it comes to workloads AI is one more of them may be a more complicated one and we in as the infrastructure folks in this industry need to learn how to serve them and as a result bringing the people closer together the AI story is a data story so the better we can manage data pipelines which I think our project Cube flow for example does really well uh the more we can understand our AI engineering counterparts the better the outcomes will be for the features and products that we deliver into people's hands yeah I think for myself cuon Paris was amazing and I think what I love about the Keynotes and themes that carried over Alex is a lot of what priyanka's mentioned um we we've seen a lot of change that's happened with the genor generative Ai and how it's affecting our community and yet here is kubernetes is still being a key Schuler that is supporting these workloads one thing that I really took away and this is coming from my end user Persona you know I always look to cucon for early signal I know I'm also sharing some of the things that we're doing at Adobe uh but it was great because then you really got some insights on just where things were at Nvidia gave some really good breakdown of what you can do today as well as tomorrow so going into Salt Lake City I'm really excited to kind of see like how did people take some of this information that came directly and I know how it's imp imped us at Adobe other areas that I think were really interesting that were really starting to develop there uh was even how do we start to digest these things in our Cloud native infrastructure and you saw some early indications of like hey this is the way that you possibly could start to incorporate here's where you would start at if you want to introduce this into your infrastructure so I found that it was very powerful and if you look at kind of what's happening in the cloud inter compute foundation and a lot of our communities it's very reflective like things with platform engineering if you go on work group platforms tremendous energy there the AI Community really stepped up rapidly not only did the churn out papers that were related to what you could consume and have a new landscape that was added for AI but then you're also kind of seeing other projects start to emerge so Salt Lake City I think is really going to just be a continuation of this amazing story that was shared there so how will this get reflected At cubec Con this year what are some of the emerging trends that we'll see in the programming prianka yes so at cucon Paris we started a a trend of doing themes for each day at cucon for the three days and we're continuing that in Salt Lake City and the first day's theme in my opinion is very representative of what we see is happening in the ecosystem the theme is platform engineering and building AI platforms so the key thing to notice here is that we are we have both of these stories on the same day there's a reason for it which is that the platform engineering teams are the ones that are being given the call to action to come build the AI platforms for companies for end users such as Adobe and uh we think that's really important uh information that our attendees need to learn from and so you'll see that reflected in the first day we have some very heavy hitter speakers I'm very proud to say who reflect by the way this exact intersection that I'm talking about so we have Ken Goldberg who so many folks and Cloud native know she was uh the leader for engineering uh uh at Google for kubernetes and now she's SVP of engineering at corvive which is a big player in infrastructure for AI uh she and their CTO are going to be talking together we also have aera Sinha another heavyweight OG from kubernetes who is now building the AI platform over at Capital One keynoting and several other awesome folks so I particularly call out these two women because they are emblematic of the change we are going to see come about in how AI workloads are run they're run by people with deep expertise and experience in distributed systems and infrastructure and we're here for that in the subsequent days the second day is The Perennial perennially important topic of security and the third day is a final hurrah for the 10 years of kubernetes what are you gonna do on that day oh so much fun there can I I don't know Joseph can I even tell them about that really fun game thing you're the coach me yeah you can I think you can probably hint at that fun because that is that is a pretty exciting closure for the tenure right so we're having a family feud Style Show which Tim Hawkin is running so I think that one's gonna be pretty special that'll be a lot of fun um so what trends are you watching in platform engineering Joseph well platform engineering has been exploded and it's really exciting to kind of see and you're seeing where you know we're we're really evolving quite a bit with it previously if you look at like the last Cube conon or even the one before that we really started to see the platform engineering maturity model like it published the concept of platform as product really became strong that we're really focusing on you know like doing research and Frameworks and guidelines that help emphasize the user needs and experience which is very important so that we're really kind of like making sure that Persona is able to kind of remove some of that friction now what am I looking forward to well I see what in the community the things that are really being talked about and we're seeing it in the in the conference tracks is areas where like people are starting to prototype these platform things where they're showcasing some of the best practices and guiding platform implementation for various use cases you see projects like canoe the backstack so we're not just saying here's some of the cloud native products we're actually showing you the way which is helping others to be able to contribute to this capability domain is being defined so that you know we can really understand the language that we all speak why we were doing this is so that when we talk about apis security observability having that standardized terminology is really going to help us to continue that collaboration in the platform engineering space and so if I had to say the trends I'm seeing there it's really that user Centric standardized scalable platform solutions that are coming out of this working group that really speaks to the maturity of what's Happening Here so what's the intersection then of of platform engineering kubernetes Cloud native Technologies and generative AI well I think that's an area where if you looked at like Paris I think there was a couple interesting things that I took away from there that I'm looking here in Salt Lake City as well we realize that a lot of it's at scale like there is some challenges with like how do I manage all this this infrastructure in an efficient way I I would say the parallel analogy I would kind of look at is what is one of the thing that's really being disruptive today in just generative Ai and and it's like coders look look what's happening there it's improving developer velocity developer productivity it's not replacing people today but really what it's doing is allowing you to do better work it's allowing you to get more done and so the same thing I'm seeing is that can we do the same thing with our infrastructure can we make it easier to find signals in observability can we do it so that we can actually manage more at scale and so when I look at what into it was discussing in Paris that was some of the early things that they were like talking about could we use AI to help make the jobs and roles of individuals who are SRE and devops more efficient less toil and just be able to really kind of have more insights and being able to provide more value to their business so I look at this intersection of really allowing us to kind of get something in a similar way with like these code editors I see the same thing going to happen in infrastructure and we'll see the introduction across the board in our platforms has this thematic breakdown Priyanka also uh being carried over to any other parts of the conference so the Keynotes will definitely represent this and then a lot of the ancillary programming that we do so for this for example this time around we are hosting table topics for lunches and the table topics each day will be based on these themes and I think these themes have been really um a sanity driver for all of us Pro Organizing the conference because as you're making decisions for what goes in a certain day versus not you can lean back on these and then make make the choices that you're going to make so you'll see it infused through and through by no means is are these teams exclusive thems it's not like you're not going to hear a peep about security or a peep about observability for that matter on you know on different days it's just a seeming so that we have a have a guidance for folks to have a narrative to follow so Joseph what kind of community aspects of cuon do you look forward to and what do you expect these themes will help with and how will you continue to Foster that growth in the community your that's become so such a Hallmark of the cubec con world so I think that's the one thing that is great about cucon is that it is really the community that has made it what it is today it's all this great content and the cncf has done amazing with it but you know if you asked me like if outside of my coacher responsibilities the one thing I love to go to is I love I love the Sig updates you know I love to really get into the core of what's Happening into those projects because right there you're really directly faced to face with like a lck of aut maintainers but one thing that I would encourage everyone to look at is there's a good talk at cucon that's related to our maintainers some of the burdens that they carry I know we talk about this a bit it's important for us to kind of look at I know recently there's been some hot takes around you know you know open source and you know it's a marketing fail or whatever but really to be honest there's a lot of people doing a lot of heavy lifting and so I think what's important is you know Finding ways to kind of learn more about how you can get more involved in the the conference in the communities and I would say dig into some of these sigs dig into what I think uh you know uh you know prianka was mentioning uh we have one thing I think that's a Hidden Gem that we started in Paris and that is with George Castro doing these great tours of our project Pavilion and it's just he just breaks it down and allows you to kind of get meet these maintainers and really get to know what's happening behind the scenes and I know from us at Adobe that's been really a strategic success for us is the more closely we've aligned with our projects the better outcomes we've had with being able to one drive back some of the problem spaces we're finding helping the maintainers understand it finding ways where we can actually contribute as well and that has been this really interesting Dynamic for us but I encourage others to kind of take that thing and and leverage some of these these areas where they can also benefit and draft off and then eventually also potentially be contributing and an active contributor in the community 100% agree with what Joseph said the project Pavilion is just a hoot so go just for the fun of it George C asro has become the most exciting Pavilion leader so to speak he take he does project tours which are so fun and most they really pull people in so don't worry if you don't know the details of this project or that project that's why you're there you're going to meet people and learn more and I think to Joseph's points that one is a Musto in your uh in your checklist I think another element to know is uh about the technical Advisory Board the end user technical Advisory Board is a new body that is of equal weight to the governing board and the technical oversight Committee in cncf Joseph is actually a member of that too and I'm very excited for the announcements they're going to make at cubec con as well because this is a group of end users who are taking leadership to talk about what reference architectures work what is the success story of let's say generative AI features in infrastructure and they are putting together a voice that can speak cohesively for the end user Community uh and what I found is I have a lot of conversations with end users of um different at different stages of their journey and the lot of questions that they ask they are going to be reflected with answers in the initiatives the tab is taking so if you're an end User it's in your best interest to go introduce yourself learn what this group is doing so that you can get on their radar for what your organization's needs are this body is just for you y that's that's great that you brought that up I totally forgot to mention that I'm part of that board we have some artifacts that I think a lot of end users are going to find very interesting I I really appreciate the cncf kind of shepherding with this group one thing that we are doing that I will H like share is that we will be publishing out reference architectures and that one thing that cncf has been very clear about is they're not a kingmaker with a lot of these project but there's some pragmatism with this as an user we do sometimes use closed Source projects open source projects and we want to really be able to kind of provide a really clear transparent like overview of some of the things that we use and so in some of these reference architectures it is really where the rubber hits the road like we're sharing how we're getting these things done so we really want to be able to kind of uplevel and not just be yes I work for a large scale company but I I was able to kind of benefit from the open source Community this is one way as an end user I feel that we strongly can be able to help other other groups depending on what size they are they will get benefits from it so really excited to be able to kind of bring this to Market here at cubec con Salt Lake City so for end users they you know for a while um there was like a group of people who were like the end user group right they were like and they were almost like ambassadors for the end User Group um I don't see that as much anymore um so I'm curious on how you're going to like is that you I guess part of the reference architecture effort is to do that but are you gonna try to you know give these people a higher profile so you can start to find more ways to get them into the community so the end user tab is as I said equal weightage to the technical oversight committee the governing board and you're going to see them on the keynote stage these folks are the official Representatives so to speak and that doesn't mean the other end user activities are stopping the user groups that you're referencing they continue to exist actually there's the research User Group which has actually done a ton of work around uh batch workloads and training AI training workloads on kubernetes there is um you know there's been Financial Services there's Transportation all kinds but these are Clos groups because they have to be so that the folks can come together and comfortably have conversations the tab on the other hand can be externally facing that is the role and so uh you'll hear a lot from them in the coming month year years and uh I have to say especially the reference architectures that um Joseph mentioned I've been in this job four and a half years I've been in this community uh nine years as long as it's been around people have been asking for reference architectures from cncf since since time IM Memorial and finally we found a way to provide the guidance without being King makers and so this I think will be a huge impact another thing that is uh continuing on is the cncf end user Radars they were very impactful in my opinion yeah oh you bringing those back yes they are coming back bigger and better okay so they'll be a little different than what they were before but and uh right now I think there's a survey out for folks to take to inform the design of The Radars but they're coming back and so all our energy is pushing you know we were in a lot of like build phase I would say on land user program for the in the last year and now the fruit it's bearing fruit and you'll see that in Salt Lake City yeah I think that what definitely from an end us perspective Alec I was going to say one thing I would encourage like end users like well first Taylor dozol has always been this really great unofficial ambassador to end users he's always seeking us out his his I I always say if you're an end User make sure you check his keynote once again it's going to be something that's going to really speak to our end users the other thing is pan mentioned kind of a few of the things but it's interesting to watch some of our working groups you know like working group platforms for example where they're actually looking to interview end users you're seeing the same thing with like our AI groups where like we're taking input directly like we're really trying to get Clos to the signal and one of the projects that I'm working on in the end user technal Advisory Board is really trying to make it easier to get input into projects from end users so they don't have to navigate all the 10 20 30 projects that they may be working with could we kind of make that a little bit more simplified more consistent and we're working closely with some of the you know major project names to be able to kind of pilot this but it's with a goal to just really kind of like close the gap and just make it a lot simpler for them to be provide their input and feel heard so just for a last question I believe kubernetes has become boring I think it's pretty evident in the apis for example um but with that boredom comes the question about what's the future of uh of a boring infrastructure you know and so I'm curious on your thoughts on that each of you sure so you know a year ago or a year and a half ago I would say kubernetes has stabilized gotten boring and uh you know it is it's having its Linux moment it continues to be those things but there's a new Fizz of energy I would say and that's come from these large-scale AI workloads the community has really stepped up to find ways to make the infrastructure story cohesive with kubernetes across all kinds of workloads that include AI workloads so there's a ton of work going on with in um working group batch there's a ton of work going going on amongst people who are trying to uh make training workloads that run on slurm work on kubernetes there is all kinds of efforts going on and so these days things are not feeling very boring to me at the same time we can enjoy the stability of kubernetes and that our distributed systems at scale are working products are shipping so it's a very very good moment in time to be in yeah I'm with panca I I I actually gave a talk last year at one of our pre-day and it was a separate project but it was it was it was isoil and something that we used you know celum and basically I declared them boring and I wanted them to be boring in my world I don't want all the you know I want innovation but I I love where kubernetes is at it is really delivering business value and for us that is important but as far as like the things like pan mentioned yeah we we're there's a lot of exciting things I'm really excited to finally see where we feel very close with things like wum is starting to become really a thing in our infrastructure just seeing now slowly serverless is starting to come emerge back and so some of these things are old are becoming new again but yet in these new innovative ways and being able to really solve some of our really challenging business problems s and then our old tried andrue security which we're going to have you know on the second day seeing some of these things where the friction is starting to be reduced down we're starting to get really clear signal like those are things that they're boring but yet to me they're really exciting they're the things that keep your eye on to be able to really bring back and you'll be able to now here we get all this great exciting things announce of cubec con but now you're starting to see it emerge into varying infrastructure you know environments so to me that is great it's a great space to be in and as you mentioned earlier like yes you know this this this is boring but kubernetes API like if you think about it you're seeing all these if you want to really see some of the emerging things there's an emerging track that I would say go take a look at because it's challenging even as I was reviewing some of this early on selecting content it was challenging even for me to think like wow people are doing this with the KU API or they're extending it with projects like kcp or you're watching crossplane and it is exciting to kind of see so it may be considered boring but there's a lot of exciting under the hood I just to add to that I think you know kubernetes is a representation of the people that have built it it's the community right and the community right now we are fighting in a good way because we are fighting to ensure that this way of delivering products in the hands of our users stays resilient stays reliable stays something that has business value and I think there are so many narrative out there at the moment of like there's there's this other way of doing infrastructure we will like totally have to build things in a different manner now and this community is saying no we have to build on our successes stand on the shoulders of giants that have come before and I think that gives this community the energy that it thrives best in which is when we're trying to demonstrate a success and show how people can work together effectively and collaboratively so we're back in that moment and it's great great well thank you both for your time and we'll see you in Salt Lake see you then if you like this video please give us a thumbs up and if you'd like to see more videos like this you can always subscribe to our YouTube channel we're on all the major social media platforms you can always find us at the new stack. we hope to see you soon [Music]
Original Description
Platform engineering will be a key focus at KubeCon this year, with a special emphasis on AI platforms. Priyanka Sharma, executive director of the Linux Foundation, highlighted the convergence of platform engineering and AI during an interview on The New Stack Makers with Adobe’s Joseph Sandoval. KubeCon will feature talks from experts like Chen Goldberg of CoreWeave and Aparna Sinha of CapitalOne, showcasing how AI workloads will transform platform operations.
Sandoval emphasized the growing maturity of platform engineering over the past two to three years, now centered on addressing user needs. He also discussed Adobe's collaboration on CNOE, an open-source initiative for internal developer platforms. The intersection of platform engineering, Kubernetes, cloud-native technologies, and AI raises questions about scaling infrastructure management with AI, potentially improving efficiency and reducing toil for roles like SRE and DevOps.
Sharma noted that reference architectures, long requested by the CNCF community, will be highlighted at the event, guiding users without dictating solutions.
Here's the full article to go with the video: https://thenewstack.io/platform-engineering-rules-now-with-ai/
Learn more from The New Stack about Kubernetes:
Cloud Native Networking as Kubernetes Starts Its Second Decade
https://thenewstack.io/cloud-native-networking-as-kubernetes-starts-its-second-decade/
Primer: How Kubernetes Came to Be, What It Is, and Why You Should Care
https://thenewstack.io/primer-how-kubernetes-came-to-be-what-it-is-and-why-you-should-care/
How Cloud Foundry Has Evolved With Kubernetes
https://thenewstack.io/the-evolution-of-cloud-foundry-with-kubernetes/
Join our community of newsletter subscribers to stay on top of the news and at the top of your game.
game. https://thenewstack.io/newsletter/
Watch on YouTube ↗
(saves to browser)
Sign in to unlock AI tutor explanation · ⚡30
Playlist
Uploads from The New Stack · The New Stack · 0 of 60
← Previous
Next →
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
What's Next for the Cloud Foundry Foundation in 2017 with Executive Director Abby Kearns
The New Stack
How Unikernels Can Better Defend against DDoS Attacks
The New Stack
Weaveworks is Bringing Horizontal Scaling to Prometheus
The New Stack
TNS Analysts Thanksgiving Special: The Evolution of Kubernetes and the Container Ecosystem
The New Stack
How Rancher Labs is Seeing Kubernetes Put to Work in Production
The New Stack
SAP Tests Kubernetes for Cloud-Native Enterprise Software Deployments
The New Stack
Event Marketing for Today's Developer Evangelists and Community Managers
The New Stack
NodeSource Introduces Certified Modules to Improve Node.js Security
The New Stack
How Lightstep is Illuminating the Case for Distributed Tracing
The New Stack
How OpenStack Aims to be More Inclusive without being Exclusive
The New Stack
How Shuttlecloud Saves Time and Money by Monitoring with Prometheus
The New Stack
Creating Analytics-Driven Solutions for Operational Visibility
The New Stack
Understanding the Application Pattern for Effective Monitoring
The New Stack
Building On Docker's Native Monitoring Functionality
The New Stack
The Importance of Having Visibility Into Containers
The New Stack
How Getting Your Project in the CNCF Just Got Easier
The New Stack
Tectonic Summit Pancake Breakfast: How to Sell Kubernetes to the Hypervisor-Minded
The New Stack
The Buzz at Tectonic Summit 2016 in New York City
The New Stack
Bringing Clarity to the Future of Node.js Modules
The New Stack
How FluentD Can Help Monitor Microservice Architectures Through Unified Logging
The New Stack
Reshaping Front End Development with Warehouse.ai
The New Stack
2016 Year End Wrap-Up: Discussing Docker, OpenStack, and Open Source
The New Stack
Here's Why You Should Build a Robot Using Node.JS: Because You Can
The New Stack
How the Node.js Foundation is Utilizing Participatory Governance Models
The New Stack
Set Up an MongoDB Replica Set in Less Than an Hour Using Bitnami Packages
The New Stack
Determining Who Bears the Burden of Ensuring NPM Module Security
The New Stack
How Intel Snap uses Telemetry and Kubernetes to Drive Enterprise Efficiency
The New Stack
How the NFL Scored a Touchdown with its Open Source React Framework Wildcat
The New Stack
Aporeto CEO Dimitri Stiliadis: When it Comes to Security, Context is King
The New Stack
The Buzz at Node.JS Interactive
The New Stack
Why Going Serverless Doesn't Mean 'No Ops'
The New Stack
How Node.js is Transforming Today's Enterprises
The New Stack
JJ Asghar Interview
The New Stack
How Capital One is Using APIs to Streamline Auto Financing
The New Stack
SXSW 2017: How Machine Learning Differs From Regular Programming
The New Stack
SXSW 2017: Data-Driven Applications with Capital One DevExchange's Hydrograph
The New Stack
SXSW 2017: How Good Engineers Make Bad Business Decisions
The New Stack
CloudNativeCon & KubeCon EU Pancake Breakfast 2017: Kubernetes and the Multi-Cloud
The New Stack
CNCF Executive Director Dan Kohn: What's Next for CNCF in 2017
The New Stack
Exploring the Latest Container Runtime Projects in the CNCF
The New Stack
Exploring the Future of the Kubernetes Ecosystem
The New Stack
Kubernetes and Continuous Deployment
The New Stack
Kris Nova of Deis at CouldNativecon/Kubecon in Berlin
The New Stack
Docker's Quest for Simplicity with the Evolution of Containerd
The New Stack
Developers First: The Cloud Foundry Service Broker API and Kubernetes
The New Stack
Mapping the Future of CoreOS's rkt in the CNCF
The New Stack
Red Hat and Dell EMC: Two Perspectives from DockerCon
The New Stack
Capital One Opened its APIs to Third-Party Developers — Here’s What They Learned
The New Stack
SUSE Joins the CNCF, Brings Kubernetes to OpenStack Cloud 7
The New Stack
How Capital One Brings Open Source To The Banking Industry
The New Stack
OSCON Is Coming Back To Portland, A Show Wrapup With Co-Chair Kelsey Hightower
The New Stack
Dev Or Ops Doesn’t Matter, You Need Observability
The New Stack
Taking The Next Steps In Developing An Open Source Culture
The New Stack
SXSW 2017: How Capital One Became Technology-First With Open Source
The New Stack
Apcera Old Apps Spanning New Clouds
The New Stack
Provenance: The Peace of Mind Chef Habitat Seeks to Deliver
The New Stack
InSpec: Human Readable, Automated Compliance
The New Stack
The Evolution of SAP HANA Express
The New Stack
Women Engineers Who Inspire And Never Give Up
The New Stack
Three Perspectives on the Evolution of Container Security
The New Stack
More on: AI Systems Design
View skill →Related Reads
📰
📰
📰
📰
How I built my own Railway at just just $2/mo with 4 CPU cores and 7.7 GB of RAM; INSANE!
Dev.to AI
Reverse Proxy
Dev.to · Gouranga Das Samrat
In Pursuit of the Ideal Developer Experience
Dev.to · Jonatan Lampa
Why AWS CodePipeline + ECS falls short for production-grade microservices (and how EKS fixes it)
Dev.to · Arnab Adhikary
🎓
Tutor Explanation
DeepCamp AI