---
title: "Customer-led Innovation | SHIFT on-demand | Commvault"
type: "WebPage"
language: "en-US"
url: "https://www.commvault.com/resources/video/customer-led-innovation"
date: "2026-11-16T12:17:29-05:00"
modified: "2026-11-16T12:17:29-05:00"
description: "Watch how Atlassian, Google Cloud, and Commvault collaborated to develop scalable Google Cloud backup and recovery with Clumio CloudRecover."
image: "https://www.commvault.com/wp-content/uploads/2026/11/Customer-Led-Innovation.png"
breadcrumbs:
  - name: "Commvault Home"
    url: "https://www.commvault.com/"
  - name: "Resources"
    url: "https://www.commvault.com/resources"
  - name: "Video"
    url: "https://www.commvault.com/resources/video"
  - name: "Customer led innovation"
---

# Customer-led Innovation • SHIFT On-Demand 2026

- [Home](/)
- Videos
- Customer-led Innovation • SHIFT On-Demand 2026

ON-DEMAND SESSION

## Customer-led Innovation: The Story Behind Bringing Commvault Clumio to Google Cloud

See how Atlassian, Google Cloud, and Commvault turned a customer need into scalable Google Cloud protection with isolated backups, consistent operations, and validated recovery.

[Request a demo](/request-demo)

Play Video      ![Video thumbnail](/wp-content/uploads/2026/11/Customer-Led-Innovation.png)

---

- [Key Takeaways](#key-takeaways)
- [About This Session](#about)
- [Related Resources](#resources)
- [FAQs](#FAQs)

[Back to all on-demand sessions](/shift#watch-now)

---

## Key Takeaways

**Customer Needs Drive Innovation**

Atlassian’s Google Cloud requirements accelerated development of scalable Google Cloud Storage protection.

**Protect Data at Scale**

Isolated, immutable backups support petabyte-scale data and billions of cloud objects.

**Standardize Multicloud Operations**

Consistent APIs and workflows simplify protection across AWS and Google Cloud.

**Validate Recovery Continuously**

Regular monitoring and restore validation help teams verify recovery objectives at scale.

**Co-Develop for Speed**

Shared code, Terraform modules, and technical collaboration accelerated delivery against tight deadlines.

**Durability Is Not Resilience**

Independent backups help recover from deletion, misuse, compromised accounts, and agent actions.

---

## About This Session

**A customer requirement became the catalyst for a joint resilience solution across Atlassian, Google Cloud, and Commvault.** Leaders from the three companies explain how Atlassian’s multicloud strategy created a need for isolated, immutable Google Cloud backups with the same operational experience and recovery standards already used for AWS.

#### **A Multicloud Requirement**

Atlassian protects about 40 petabytes in Amazon S3 and three petabytes in DynamoDB. Its environment includes buckets with billions of objects, a one-hour recovery point objective, and a six-hour recovery time objective.

**Moving core platforms to Google Cloud required equivalent protection without introducing a separate backup stack or fragmented operating model.** Atlassian’s request accelerated Commvault’s Google Cloud Storage roadmap and shaped enterprise capabilities for onboarding, automation, and restore validation.

#### **Three-Way Co-Development**

The teams met regularly, shared technical requirements, contributed Terraform modules, and worked through permissions, policies, regions, and architecture in a combined development effort.

**Google Cloud services including Eventarc, Cloud Run, and Infrastructure Manager supported distributed processing, serverless scale, and isolated environments.** A common interface and uniform restore validation help Atlassian maintain a cloud-agnostic platform with less operational complexity.

#### **Durability Needs Recovery**

Cloud durability does not prevent accidental deletion, compromised administration, or unwanted agent actions. Independent backups and tested restores provide another recovery path when native controls are not enough.

CONTINUOUS CLOUD RECOVERY

## CloudRecover

Recovery built for the cloud. Proven at scale. Ready for AI.

[Explore solution](https://www.commvault.com/solutions/cloud-recover)

CONTINUOUS AI RECOVERY

## AgentRecover

Full-stack AI recovery.

[Explore solution](https://www.commvault.com/solutions/agent-recover)

---

## Related Resources

Supported tech

## Google Cloud

Protect Google Cloud workloads with scalable cyber resilience.

[Explore the tech about Google Cloud](/supported-technologies/google)

analyst report

## Architecting Cloud Data Resilience

Build a stronger strategy for protecting cloud-native data.

[Read the report about Architecting Cloud Data Resilience](https://www.commvault.com/resources/analyst-report/architecting-cloud-data-resilience)

BLOG

## Multicloud Resilience

Simplify data protection and recovery across cloud environments.

[Read the blog about Multicloud Resilience](https://www.commvault.com/blogs/multi-cloud-recovery-that-actually-works)

---

## Frequently Asked Questions

Why did Atlassian need independent cloud backups?

Atlassian needed isolated, secure, immutable protection against ransomware, rogue administrators, compromised accounts, accidental deletion, and other scenarios where platform durability or native recovery capabilities alone would not be enough.

What scale does Atlassian protect with Commvault?

The session describes about 40 petabytes in Amazon S3 and three petabytes in DynamoDB, including S3 buckets with billions of objects. Atlassian monitors recovery objectives and regularly validates restores.

How did the Google Cloud solution get developed?

Atlassian’s requirement accelerated Commvault’s Google Cloud Storage roadmap. Atlassian, Google Cloud, and Commvault collaborated on requirements, architecture, permissions, Terraform modules, automation, onboarding, and restore validation.

What benefits come from one multicloud backup platform?

Using comparable APIs, workflows, and validation across AWS and Google Cloud reduces separate code bases and operational divergence. It also helps teams maintain a consistent resilience posture while remaining cloud agnostic.

Why is cloud durability different from resilience?

Durability protects against infrastructure loss, but it may not prevent deletion or misuse by an authorized user, system, or agent. Resilience adds independent backups, administrative controls, validation, and tested recovery paths.

---

## Transcript

View full transcript

Good morning everyone. Thanks for joining us. So today we want to spend some time on what happens when a customer challenge becomes a catalyst for a product innovation across three companies Google Cloud, Atlassian and Commvault. So before I jump in, I thought we will do a quick round of intros for all of us.

I’ll start with myself. Poojin Kumar, the Co-founder and CEO of Commvault. Hi everybody. My name is Lin Chen and I am from Atlassian. I’m the Head of Technical Program Management for Core Data Platforms at Atlassian. Hi, I’m Seema Joshi, I’m the Head of Engineering for Trusted Data platform. One of my teams is responsible for the backup and restore of all data stores at Atlassian.

Hey everyone, I’m Nick. I lead the Customer Engineering Team over at Google for software vendors headquartered at of New York. This is my first SHIFT. It’s been awesome. And thank you guys for joining this session. Awesome. So let’s start. I guess the first question is for you, Lin. Tell us about the Atlassian Cloud journey. What problem is Atlassian trying to solve and why did it partner with Commvault?

Yeah, great. As many of you probably would know that Atlassian started off as a small developer tools company in Australia. This is over 20 years ago now, and and it’s slowly evolved into a major enterprise company providing enterprise collaboration tooling, which includes our flagship products like Jira and Confluence. We started with our server products deployed in the data centers and then moved our product into the cloud.

And we moved a lot of our tenants and customers into the cloud. And as we expand it into the cloud and started supporting a lot more larger enterprise companies, our security and compliance posture basically, you know, rose significantly. So this is when we realized that we needed a truly isolated, secure and immutable backup solution to protect against attacks like ransomware and rogue admins and compromised admin accounts.

So when we started this journey, we evaluated a bunch of different backup companies in the field, including the native AWS backups. We found that none of the solutions was able to provide the scale that we needed, while meeting all of the functional requirements that we had, Commvault was the only solution that was able to help us support our large scale and support our RPO and keep our data in the isolated, immutable environment.

So maybe you want to just provide like how how big our data size is and what that scale is. Yeah, sure. So we use for both S3 backups and DynamoDB backups in AWS, and we store somewhere around 40PB in for S3 and somewhere around three petabytes for Dynamo. So it’s a smaller footprint in Dynamo. But some of our buckets in S3 have billions of objects.

So we have we do regular RPO monitoring and restore validation to test whether we can restore in time. Our RPO is currently one hour. Sorry, RPO is one hour, RTO is six hours. So was able to meet those timelines for us to restore such large tables. And this is where a partnership has been working really well for the scale.

The other thing we have is we do some selective tagging of our buckets so that we exclude them from backups. And that is another feature where Commvault helped us so that we don’t pay too much earlier deletion fees for, you know, objects which get deleted quite quickly. Awesome. I think it’s been a fantastic collaboration effort, and we’ll talk about that later.

But before that, you know, kind of fast forwarding today to to today. Lynn what’s Atlassian strategy with with Google. Yeah. So our partnership with Google has is one of our most strategic initiatives right now at Atlassian. So basically we want to run our core platforms natively in GCP. And when we do that, we are officially becoming a true multi-cloud company.

And this really enables us to bring Teamwork Collection, which includes our flagship products Jira, confluence, Rovo, and Loom. Basically, it brings all of the customers who couldn’t use our cloud before due to hard rules against using AWS or strict GCP compliance requirements, and internally. Having this initiative also is an opportunity for us to extract our platform layer and our product into a, you know, a cloud agnostic platform.

So basically, you know, we have the flexibility today to move into future clouds and future environments. So this is a really great initiative for us, for our customers as well as for us internally. That’s awesome. And then, you know, we are always in the picture. When you were in AWS, how did you reenter in getting the picture with Google?

Yeah, that’s a great question. And this is something that I personally had involvement in. So, you know, as part of our Google collaboration, one of the critical requirements is to have backups in our Google platform. The same setup as what we have with us. Again, we looked for solutions. We worked with different vendors to understand what they provided.

But you know, one thing that we wanted to to keep was the same exact stack. We didn’t want to, you know, having to have to support multiple different stacks for, for our multi-cloud solution. But the challenge here was that at the time Commvault didn’t offer a backup solution for GCS backup. So I talked with my team and we kind of said, well, why not just ask if they be interested in supporting something like this, supporting a GCS solution?

And so I went and reached out to the team and as it turns out, they were already, you know, exploring this solution. So they just really needed a customer like Atlassian to give them that conviction to go forward and deploy and launch this product. So we have we had this joint collaboration and it has been an incredible journey.

And Seema will talk a little bit about that in a in a minute. But what I can say is that we are on track to deploy our backup for GCS, for our Google launch, which is going to happen later on this year in December. Awesome.

So moving to you, Nick, again, you know, from your perspective, how did you see this customer requirement and how did this become strategically important for Google Cloud? So I manage a team that covers software vendors out of New York City, and it actually brings me so much joy when I get to see customers driving further adoption of software solutions on Google Cloud.

And also, it’s just delightful when I see software organizations like Atlassian and Commvault providing differentiated features on top of the base platform, which is Google Cloud. So across the hyperscalers, Google Cloud, Azure, AWS, our goal is to revive the infrastructure layer for software builders like yourself. Commvault to be able to provide even more value for our collective organizations.

So I just it brings me so much joy to see us with Google presenting here today in terms of things that really stuck out to me. So the two big takeaways for me. One is getting that last flagship products. Tens of thousands of customers use these flagship products globally on Google Cloud. So now any Google Cloud customer that wants to deploy Atlassian Jira Confluence on Google Cloud now has that support.

That’s takeaway one, takeaway two is coming. And is creating a blueprint for how to actually build this solution on top of Google Cloud to provide true resilience on top of Google Cloud storage. So critically important for Google, critically important for our team and really appreciate the partnership that we have together. Awesome. And I’ll tell you, I think this collaboration effort has been remarkable, right?

Between, you know, Google and Atlassian and coming out right. We you know, if you go back into our journey, you know, many years ago, we started on AWS and we went very deep in working on AWS. But then when we and we worked very closely with, with all the, the folks on the AWS side, but then, you know, moving into the Google Cloud, it wasn’t as easy, like, oh, we do whatever we did with AWS.

You’re now basically just ported into Google Cloud. It doesn’t work like that. You have to go and think through, you know, how does the permissions work. You know, how does the architecture work with with Google and how do all these things in a fit in? And then again, you know, doing that while making sure that we are servicing the requirements from Atlassian.

And that’s where you’re working very closely with that Commvault team. Our, you know, our team and our architects along with their folks, so that we essentially are making sure that we are giving them the regions and we are making sure we are doing the right policies and permissions. That is is the right thing from Atlassian perspective also. And this time, I would say the collaboration also went and we could do we could execute much faster and we could do things much faster because Atlassian not just provided input, but they also provided, you know, and worked with us in terms of collaborating with code and Terraform modules.

And then and they did all of that. And it’s almost like a joint development effort that we did that basically led us to supporting, you know, their requirements at a rapid pace. And from a Google perspective, again, we’ve seen the same collaboration that that we saw before with working with other hyperscalers, because we were able to work closely with their technical teams and essentially get get the expertise from their side so that we could essentially go and really understand the new answers of what it would take to kind of, you know, build this platform on Google Cloud.

So again, this is, you know, through and through was a great collaboration effort from me. Obviously, I think this time, as I said, we went to the next level in terms of even collaborating code from your side. And our teams have worked very closely in a combined Slack channel for for many years.

Tell me from your vantage point of view. How did this whole thing come together from R&D perspective? Yeah, sure. So this has been a long journey. We first reached out to Commvault about 18 months ago on our GCP request, like Lynn was mentioning, and, was already thinking about entering GCP and this was just an accelerant for them to put that on the roadmap.

We met regularly. It was a very agile process, to go over the requirements and figure out what would be delivered and what timeline. We were early access users, so we got access to the product in March of 2026. And in July, I think you guys were in GA. So we were able to provide more details on the requirements.

So some of our enterprise requirements were to have automated onboarding and restore validation for the scale. The goal was to have the same resiliency and operational, you know, posture as AWS. So we wanted integration with our identity module and our identity infrastructure adopting Terraform. Like you mentioned for provisioning. And this is where my team collaborated with Commvault and kind of made sure that it integrated.

Well, like you mentioning Poojin, it was a long journey because, Atlassian has been AWS native. When we first went to cloud, we went AWS native. And so we had to migrate to cloud. So as part of that, we moved to Kubernetes for compute. We went to Kafka for streaming, and then we were looking at options to support all of that.

So it was not a simple lift and shift, but at the same time we would partner together and be very agile with us to make sure it landed on time and was on track. Now, I think that is exactly how we have mimicked the same thing right from our side. You know, is that last scene has gone to be, you know, multi-cloud and kind of abstract these things out.

I think this collaboration has also helped us to make sure that as we have gone in the multi-cloud journey, we have abstracted everything out so that we can basically deliver the uniform experience to our end customers. Nick, from your perspective, you know, tell me your thoughts in terms of this, this collaboration from a Google Cloud perspective. I think you’re both heading on a lot of the topics from the Google side.

I think it just brings me so much joy when I see sort of organizations pushing our teams to be better. So within the software vendor space, there are a lot of very technically deep individuals who work on the engineering teams there, and seeing the feature requests, seeing the quality of the technical sessions, which we held together. It really just showed how both your organizations have a very high bar when it comes to technology.

That actually helps push our overlay teams, our product teams, to understand what you need to build effective software to deliver the business outcomes which organizations are looking for. And throughout this journey, I can recall many situations where we were sort of in a war room in a deep technical conversation, debating and discussing how to achieve what we’re trying to accomplish based off of sort of the current state of Google Cloud services relative to what you’ve experienced from other providers, and also based on what you were trying to accomplish for Atlassian.

So I think we landed on a really great architecture. On the Google side, we leverage Eventarc for sort of distributed message busing technologies on the side. We landed on Cloud Run for serverless compute to scale out the sort of backup and resiliency jobs, and then also in terms of being able to create isolated environments from a resource and project perspective, landed on a Terraform wrap tool called Infrastructure Manager from the Google cloud side.

So I think the commentary around not just going copy paste from the previous architecture from the other place. I think that that brings me a lot of joy. So I’d say multi-core effort across three huge organizations, great outcomes across the board. And just it’s been delightful to partner with you both. And I just want to add to, you know, you said the three companies, it’s hard to collaborate with all of that.

And we were running against a really tight deadline for our launch. So it was awesome to see that, you know, the collaboration just happened so quickly and so smoothly. And we’re able to, you know, launch our product on time with all three collaborators. So it’s awesome. No, absolutely. I think, you know, there was a time when, you know, the Google cloud migration was basically kind of almost contingent on, you know, just their work internally, but also the support that we have to have for GCP.

So again, there was a lot of, you know, time pressure in the system. So which is where I think, you know, without this collaboration, I don’t think we would have achieved this in the time frame we did. Again, you know, coming back to this, you know, what benefits are you seeing with this and how does this, you know, advance your own broader resilience goals and cloud goals?

Yeah. Like when we initially started, we did not have a solution for backup on GCS. And that was where Commvault was a key enabler for us to meet our our goals. And we do have the same interface with our AWS solution as we have a GCP now, because we use the same solution or use across the board. So the APIs are consistent for us.

It means better maintenance for my teams to interact with the same kind of interface. And I think that was a conscious decision so that we could have that abstraction and we could be cloud agnostic in a way that you don’t have different code bases and you don’t diverge too much. We put in a lot of effort as well to have uniform, restore validation across the board as well.

So this is where having Commvault as a solution for both AWS and GCP was beneficial for us. It’s a natural extension for Atlassian to partner with Commvault on GCP as well. And again, we’re super excited. Again, we have we have we have done this partnership. And Atlassian has co-developed build on top of our, you know, Terraform and SDKs and all of that.

And now if you heard the keynote yesterday, we are also launching our MCP server and kind of going and building more functionality into the product. So, obviously super excited about, you know, Atlassian and kind of adopting some of these things that we are doing and simplifying lives for, for you and your team.

And also now we have more threat capabilities with the with the ThreatDetect announcement we also made yesterday. And so again you know, that’s another opportunity where you know we can hopefully, you know help you guys to advance your cyber resilience goals. Nick coming back to you right. How do you see this from a from a Google perspective?

How does it help your customers in their cyber resilience get more and more enterprise customers on on Google Cloud? So I mentioned that during my introduction, this is my first SHIFT and this conference has been wonderful. I think just being here has really educated me more around what resilience operations truly means.

And hearing from the keynote speakers, hearing from the subject matter experts. I think that when it comes to resilience, if you were to ask everyone in this room how you define that, we’d probably get a variety of different answers in terms of what the definition actually is. So I think, when you ask that question, how is resilience so critical for cloud transformations today?

I think there’s a few key industry trends that are happening. One, there’s a larger sort of architectural pattern known as agents and AI, which brings a whole set of different endpoints which may be protected to ensure that there are backups and ability to recover when it comes to resilience for agentic systems, that in itself has built a lot of different data stores, like this concept of more usage of vector databases.

Also, there’s kind of the runtime side on the agent end and understanding context and session memory, like all of that has been bolstered into understanding how you can be resilient if you have a failure, if you have downtime, how do I bring back that a system backup? So that has kind of emerged and developed really over the last year or two.

I’d say another thing here is there are organizations that have different relationships with their resiliency strategy. There are cloud native organizations that were born in the cloud that deployed as a service and sort of have built an understanding of how to build available systems on cloud there were born, let’s say, decades ago, who started in the data centers, who understand physical machines, and what that meant in terms of having a business continuity disaster recovery solution and also the availability properties of that architecture relative to cloud.

There’s certain nuances there. So I think that for cloud transformation, it’s really important to understand where you are within that sort of customer lifecycle, understand some of those industry trends, and then be able to really build out requirements around what is important for us to be able to have resilience across our platform. I think that an important thing within cloud environment is to understand it’s not just about the backup and the copy across a different region.

It’s also important to think about the administrative controls. In case a user or a system or an agent goes ahead and does something that it should not do. So you might have the best architecture, but you could have a system that deletes an object and not big storage. You have an agent that drops a table in a sort of business critical database, right?

And having tools like ensure that you have that sort of full encompassed resiliency wrapped around these core systems and the architecture. Distributed systems are becoming way, way larger. So with MCP and AI and all these new data stores that support these core workloads, having a cohesive sort of resilience strategies, very important for that cloud formation, what you’re calling out.

Yeah. And totally echo that, right. I think, you know, people confuse you know, durability with resilience. Right. And so durability is there with the cloud. You know, they provide you durability. But then that doesn’t prevent you from, you know, going and, you know, accidentally or otherwise, you know, deleting stuff. So which is where, you know, as we have seen, you know, in our journey over the last few years with AWS, we see that exactly translating into our support on GCP, because the scale, the simplicity and all of that that we have basically built, you know, with this collaboration and taken all the goodness that we had and gotten now into, into GCP, that will help enterprise customers that are now coming on the GCP platform. Because while Atlassian obviously has massive scale and they have they started on this journey on the cloud many years ago, there’s a lot of customers who are just, you know, starting today or in different phases of their journey in the cloud, but the scale requirements are essentially going to come.

And more so now with applications being built in the cloud. And these are cloud native nature, right. They they use the same substrate in terms of, you know, Google Cloud Storage and, and Bigtable and so on and so forth. And so we go and support these things at at massive scale. And now we are only providing more and more cyber capabilities on top of the platform.

I see, you know, this just being the beginning of the journey where we are going to have, you know, like Atlassian, you know, many more customers that are going to basically on, on Google Cloud with, with the, with the same sort of set of requirements that we can now essentially go in and service. So this collaboration I think has been a phenomenal thing for all of those customers that are essentially going to go and benefit from.

So thank you very much on this combined work, both Atlassian and Google Cloud and really appreciate this. And just to add, I mean, we have exercised the restore for real customer situations and hot situations a few times at Atlassian. So, it does come up that, you know, we have PITR even for DynamoDB. Do we still need a backup?

And the answer is absolutely yes, because there are scenarios where you need to tap on this situation and restore it from backup. So I think that investment is definitely where we continue to see value for our customers. Yeah, I think exactly. I think and you know, the scale at which, you know, these guys operate as, you know, unprecedented, right?

You know, the first times you guys did it like, oh we were like up at night. Now we sleep peacefully at night because I think we have solved all the all those problems. But but again and this, this has been really, really phenomenal. Any other closing thoughts from any of you? I think we are we are happy with this collaboration and we will continue to keep using Commvault and for our backups.

Yeah, absolutely. It’s been it’s been amazing. And it’s it’s really great to there’s like kind of no boundaries in this collaboration. We, you know, we are talking to at least to, team as if it’s, you know, part of our team, like, hey, we want this feature. What do you think? And and it’s so it’s almost like we’re a part of this, this group collaboration together.

And, and it makes things go faster and it makes, you know, things go, better in this way. In order to mention our combined Slack channel, as ever, increasing number of developers between the between the two companies and now three companies. But again, really appreciate this for anyone in the audience that’s evaluating sort of solutions around resilience on Google Cloud.

Feel free to reach out and a lot more Google Cloud supported endpoints coming. And yeah, it’s been great to be the sort of hyperscale to provide this infrastructure for you all. Yeah. Hey, who better than Google Cloud to announce a roadmap? And again, thank you very much, all of you. And again, we will be around for a little bit if you have any other questions for us.

But really appreciate your time and thank you very much.
