Red Hat Summit 2026 | Venkat Ramakrishnan, NeuBird AI

Clips
More from Red Hat Summit 2026

Venkat Ramakrishnan

President and COO

NeuBird AI

play_circle_outline NeuBird mission: eliminate downtime by reducing production cognitive overload.

play_circle_outline In-Place AI Causal Engine That Correlates 30–100 Telemetry Sources for Fast Root-Cause Detection Without Data Ingestion

play_circle_outline Reduces SRE toil, frees engineers for higher-value platform work.

play_circle_outline Partnering with OpenShift at Red Hat Summit: Tackling Infrastructure Complexity Beyond Human Capacity

Info
Transcript

Venkat Ramakrishnan, NeuBird AI

Venkat Ramakrishnan

President and COO NeuBird AI

In this interview from Red Hat Summit 2026, Venkat Ramakrishnan, president and chief operating officer of NeuBird AI, joins theCUBE's Rebecca Knight and Rob Strechay to discuss how agentic AI is transforming production operations from reactive firefighting to proactive incident prevention. Ramakrishnan opens with a blunt diagnosis: production infrastructure has outgrown human understanding. NeuBird's answer is a causal reasoning engine that correlates signals from 30 to 100 data sources — pulling directly from metrics, events, logs and telemetry — to pinpoint... Read more

explore Keep Exploring

What problem does NeuBird AI aim to solve? add

How does NeuBird differ from other observability platforms, and how does it address alert/tool/process fatigue? add

How does a 24/7 AI production-ops agent affect SREs' workload, incident response, and an organization’s operational costs/headcount decisions? add

Why did NeuBird choose to partner with Red Hat and its OpenShift platform? add

bolt Powered by CUBE AI

Venkat Ramakrishnan, NeuBird AI

search

Rebecca Knight

>> Hello, everyone, and welcome back to the Cube's live coverage of the Red Hat Summit 2026 here in Atlanta. This is day three. I'm your host, Rebecca Knight alongside of Rob Strechay. Rob, it is day three and yet the energy in this room is still palpable.

Rob Strechay

>> It is. There's so much going on, so many great partners, and just so many ways to gain efficiency in this environment, and I just love where we're going.

Rebecca Knight

>> Indeed. Indeed. Well with that, I would like to welcome our next guest, Venkat Ramakrishnan, president and COO of NeuBird AI. Welcome.

Venkat Ramakrishnan

>> Thank you for having me, Rebecca and Rob. Always good to get in touch again and talk about all the interesting things.

Rebecca Knight

>> Yeah. NeuBird is a relatively new company. It's a focused company founded in 2023, came out of stealth in 2024. Tell our viewers who are not familiar with what you do. Why NeuBird exists?

Venkat Ramakrishnan

>> Thank you. Great question. NeuBird exists for the primary reason that production has outgrown human understanding. Where we are running application infrastructure, either on prem, OpenShift, now that we're at Red Hat Summit, or on cloud, on any other different cloud providers, the production infrastructure has become so complex, moves so fast and all the businesses are betting it, betting their businesses on the production infrastructure, it is really outgrown human understanding. So NeuBird AI exists to help platform engineers, cloud platform engineers, DevOps, SREs to reduce the cognitive overload that the production infrastructure has, gives you, and identify and resolves issues very quickly, but not just that, to prevent issues altogether. Our mission is to eliminate downtime.

Rob Strechay

>> Yeah. I mean, we've talked before, but different era, and I think, but you've always been around the cloud native aspects of this forever like we both have. And I think one of the things that I find interesting with NeuBird is really, it's going after this problem of observability. And observability, people have spent lots of money on observability, and a lot of organizations have multiple tools in there. What is different about how NeuBird is really approaching this issue?

Venkat Ramakrishnan

>> Excellent question, Rob. You always ask great questions. There's a lot of observability platforms. There's a lot of tools, right? We went from alert fatigue, right? I mean, if you look at it, there's tons of alerts, millions of alerts that come in. We went from alert fatigue and there were tools that were invented to reduce alert fatigue to tool fatigue, right? Now there's a process fatigue. So the problem was continued to become worse. It hasn't gotten any better. So now, instead of dealing with one tool, enterprises deal with tens of tools. Instead of dealing with reducing alerts, now they just put more and more rules and filters on existing alerts. And it doesn't stop. It's like the mail. It never stops coming. Alerts never stop coming. What NeuBird takes is an approach that is uniquely different from anything else out there, even the observability platforms, we can directly attach to the source of metrics, events, logs and telemetry, and we have built a causal reasoning engine on top that builds context dynamically in real time and can correlate signals from about 30 to 100 different sources and identify and pinpoint the issue within a couple of minutes. Literally, we can find the needle in the haystack within a couple of minutes, right? But we can also run on top of existing observative platforms. We run on top of any of the observability platforms that are out there today. We can pull all of the telemetry from there, but we can go directly to the source as well. We do this all without storing or ingesting customer's data because most of the platforms out there, you have to ingest your data into their platform. They'll charge you an arm and a leg and then you get some results out of it. But what Newburg does is in-place AI analytics, AI root costs, AI production ops where we never have to ingest data and it's read only. That means we don't change your infrastructure unless you ask us to, you give us permission within your guardrails. So it's a fully functional, full stack production operations AI agent.

Rebecca Knight

>> How are you building trust with SREs and engineers who've heard maybe a version of this and are skeptical because there is such a fear, a real fear that AI agents will go rogue and break stuff or make a bad call or miss the context? What are you doing to build trust there?

Venkat Ramakrishnan

>> Yeah. That's an excellent question, Rebecca. And this is a question that a lot of customers ask us also like, how can I trust your agent? The best way to build trust, the best way is to bring in an AI agent and give it the guardrails. For example, we have always said we are a read-only agent, right? And if you want us to make a modification, then we have best practices. Like for example, we'll work with your GitOps infrastructure. We integrate with your ServiceNow or Jira so we can actually create a task for your human in the loop to go complete it and you can review and approve it as an approval process. So that's number one. Number two, as I mentioned, we do not store any of your data. The biggest concern for people is that, "How much of data are you going to store? How much of it are you going to use it for your gains and purposes as opposed to, are you going to train your models on our data? Are you going to use it to build your engine?" And we don't store any of your data. We build a dynamic context in your environment. I mean, that's basically primarily for your environment. The third thing is we don't use your data to train our models. In fact, we are model agnostic. That means you can bring your own models, BYOM, and we can work with your models and we can use your ChatGPT or OpenAI or Claude or any other frontier model tokens and we can generate, deliver the same results with 94% accuracy within a couple of minutes at the lowest cost. What we have built is an agent context platform. That platform helps us surgically get to the root cause a lot faster and shrink the context window so we use the least amount of tokens. So we are token efficient as well. So with all this, when we explain this, when we talk to SREs, platform engineers, we show them the platform of trust that we have built that keeps their data to them, does not use a lot of tokens and they can bring their own models. And on top of it, we are SOC 2 compliant and certified. We have other certifications as well that is, we are compliant with all the regulations out there. So we are a fully secure platform that we can trust in deploying production.

Rob Strechay

>> I was going to say, just take that a step further about what agentic AI for operations really means because we've been talking agents all week long. So let's unpack what it means for that operations' perspective.

Venkat Ramakrishnan

>> The most important thing today is reactive operations. I mean, where an alert comes in, we can rapidly triage and identify root cause and even gen data fix and that can be applied. But I think most of the agentic enterprise ops are built for that and stop there. What we are doing is very unique. We actually have built a prevention engine. We can look at around the corners, we can look at how your infrastructure is deployed, where the blind spots are and we can start working with your engineers to how to resolve them, to give them steps or run books or even a source code. If you allow us to integrate it with your GitHub or a Copilot MCP, we can actually deliver a fix to you as a PR or to your Jira so we can show them how to avert the crisis instead of handling a crisis after it happens. So essentially, what we are trying to do is to eliminate the war rooms. Everybody else is trying to say, "Okay, I'm going to add an AI agent to war room." I mean, that's still very reactive. You're still waiting for the disasters to strike and somebody to help you. Of course, we do that. That's table stakes. We do that a lot faster than anybody else, but what agentic AI for enterprise office really means is can we actually prevent incidents? It's not a better mousetrap. It's a completely different paradigm. It's a change in mindset. And the same thing about costs. When you are looking at... When we can go to all the sources and when we can deliver a root cause pretty quickly or even predict and prevent issues, is there really a need to put a lot of data in your observability platforms? I mean, step back and think about it. Why did we need these platforms? Because we had to go back and do postmart among why an incident happened, but prevention is better than cure. If you're actually preventing an issue, is there really a need to go invest millions of dollars? And companies spend... Most global 2000 companies spend hundreds of millions of dollars on observative platforms and logging platforms and these tailgate outages, the MTTR is still pretty poor and the reliability is still trends low. So when we take a different approach... That's what we did. That is what really agentic AI for enterprise ops, what it means is it should not just deliver an incremental capability. It should change the way, change the behavior altogether.

Rebecca Knight

>> Let's talk about that for a second because you're saying it's not replacing the humans, it's not replacing the SREs, it is replacing the war room. So talk a little bit more about what you said in terms of cognitive load, that the SREs are dealing with a lot here and you are taking that off their shoulders.

Venkat Ramakrishnan

>> Exactly. See, SREs... Actually, if you look at the actual role of an SRE at a platform engineer, there's a lot of value they can build, right? They're in the process of architecting large scale infrastructure, defining how their internal customers or apps consume the infrastructure, defining what are the error budget and how they can architect within that budget. But instead, what do they end up doing is that 95% of the time, if you go to an SRE team, about half the week working week, they're dealing with fires. What we are giving them is a 24/7 AI agent that cuts out 95%, 99% of the human toil involved in dealing with production infrastructure. So that frees them up. Really, big time, most people have told us that they've gotten back at least two days back every work week that they are just doing reactive break and fix, just patching stuff up. Where we came in and started people running into production, they have said, "No, I've got a lot of time back." And now that can be applied to actually improving the service, re-architecting it, moving to a newer platform. So that time back. And for executives, it's also like how you manage your OPEX. Most executives go into a financial planning unit and say, "Oh, I've had 10,000 incidents. I need 10 more engineers to handle this." So we tell them no, instead of like just keep on adding people, get NeuBird as your production ops agent and reinvest those 10 heads into actually building a better platform, right?

Rob Strechay

>> Right.

Venkat Ramakrishnan

>> So it frees up these OpEx dollars also, along with giving people their lives back, pretty much. So that is what we see when it comes to cognitive overload or toil, toil reduction.

Rob Strechay

>> Yeah. Toil is always at the root of a lot of people who have been here, but... We're at Red Hat Summit. Again, we've been around this cloud native area for quite some time, usually talking at KubeCon and Red Hat Summit, which we've done a couple of times, but I think one of the things that why bring NeuBird to Red Hat Summit? Because there's other observability companies. There's also that are claiming some AI. There's also some others that are here that can do paging and do other stuff that are claiming some AI, but there's also things that seem to be on a parallel and complimentary in things like Ansible and some of the stuff that they're doing with AI and agents and stuff. So frame it up for the people who are here and watching from home that why NeuBird and Red Hat?

Venkat Ramakrishnan

>> Yeah. First of all, I'm a big fan of Red Hat. In my previous jobs, I think OpenShift is a very, very powerful platform. It is something that the industry needs today and it's also the platform of the future. Hybrid cloud is the future. But as powerful as it is, there's a lot of workloads, a lot of applications coming into OpenShift, but what you see is that, as I've mentioned initially, the production has outgrown human understanding, but also has outgrown human capacity. If you look at many of the teams here, they are handling tens of thousands of OpenShift worker nodes, millions of microservices. If you look at their team size, it'll maybe five or six people, and those teams are not going to grow at all because there's ton of automation, ton of tooling, and the platform engineering concept. And plus, almost every enterprise out there is trying to manage the budget and manage the OpEx. So there is a lot of pain that can be solved by a production ops agent like NeuBird. So I look at this as a great opportunity for NeuBird to partner with some of the greatest companies here and some of the innovators and like folks who run OpenShift in production at scale and this is an ecosystem that's vibrant growing fast. So one of the reasons I wanted to be here back here, even despite my previous gig, also selling to OpenShift customers. The second thing is, of course, everybody talks about AI, and I look at that as a great opportunity because now the mind share is there. When Red Hat talks about, "Oh, we want to bring in agentic operations to OpenShift," it's a validation that is needed. I don't have to go and tell people you need agentic operations because... Red Hat is telling everybody. Everybody here is talking about it. So I look at that as, look, the space is being validated by Red Hat. There is a big burgeoning need and there's a ton of customers who need a solution and we want to give them a choice. We want to be part of the journey. Plus, we're not a very opinionated solution that says you can run us with Red At, but you can also run it with other platforms like WeCenter. You can run us with other cloud providers, AWS, Azure and GCP. In fact, Microsoft's an investor in Newburg. And we are an AWS and Azure startup programs as well. So we give customers a multi-cloud agentic experience that they can bring one agent that covers all of their deployments, and most all enterprises here have at least one cloud as a partner, if not two, plus they run in data centers. So this gives them a perfect choice. They're not tying their agent to one implementation, one platform. It's a platform independent AI that can literally be a single agent, a super agent for all of their platforms. That's why we are a great choice for even customers running OpenShift either on prem or on the cloud. So that's why we are here.

Rebecca Knight

>> One final question, when we're back here next year at Summit 2027, what is the headline you want to be sharing with us about NeuBird?

Venkat Ramakrishnan

>> Well, the headline is NeuBird AI has cut the downtime by 99% for all customers running Red Hat OpenShift.

Rebecca Knight

>> Okay.

Venkat Ramakrishnan

>> The one percent is out of our control. .

Rebecca Knight

>> . Excellent. Okay. And bold, ambitious. Well, thank you so much, Venkat.

Venkat Ramakrishnan

>> Thank you.

Rebecca Knight

>> A pleasure having you on. I'm Rebecca Knight for Rob Strechay. Stay tuned for more of the Cube's live coverage of the Red Hat Summit. You're watching The Cube, the leader in enterprise tech news and analysis.