Test

AI agents sound exciting. But my conversation with A. Ravi M., CIO at Box at Google Cloud Next '26 on The Ravit Show was not about excitement.It was about risk. We are moving from AI that answers to AI that acts. And that shift introduces a completely new set of challenges. Not just accuracy, but control, access, and accountability. Ravi pointed out that most enterprises are not struggling with AI capability. They are struggling with governance. Who has access to what data, what an agent is allowed to do, and how you track those actions. Those gaps become very real once agents start operating on sensitive enterprise content.

And that is where security needs to evolve. It is no longer enough to protect data at rest. You have to think about how AI agents interact with that data in real time, and what guardrails are in place when they take action.

The partnership with Google Cloud plays a big role here. With platforms like Vertex AI and BigQuery, the focus is not just on building agents, but on building them with the right controls and visibility from day one.

The biggest takeaway for me was simple. If you are a CIO thinking about AI agents, do not start with deployment. Start with trust. Because without that, none of this scales.

#data #ai #box #security #googlecloudnext #api #google #theravitshow

Everyone is talking about AI agents, but after my conversation with Ben Kus, CTO at Box at Google Cloud Next 2026 on The Ravit Show, one thing became very clear. Agents are useless without "context". #boxpartner

Ben kept coming back to that word. Not just data, not just models, but context. In an enterprise setting, context means understanding the full picture around your data.

Who created it, where it lives, who can access it, and how it should be used. Most companies already have massive amounts of content, but it is fragmented and static, and that is the real problem.

What stood out is how Box is approaching this. They are not just storing enterprise content, they are structuring it in a way that AI agents can actually use, turning content into something agents can reason on, not just retrieve.

And this is where the partnership with Google Cloud comes in. With models like Gemini and platforms like Vertex AI, they are able to operationalize that context at scale in real workflows.

The biggest takeaway for me was simple. If you want to become AI-first with agents, do not start with the agent. Start with your data. Structure it, govern it, and make it usable. That is what actually makes AI work.

#data #ai #box #security #googlecloudnext #api #google #theravitshow

BREAKING: Kore.ai launches Artemis — a new generation Agent Platform for enterprise AI

I just sat down with Prasanna Arikala at their San Francisco office right after this launch.

And here’s what stood out.

For years, most enterprises have been stuck in the same loop:

-- Build AI pilots

-- Struggle to productionize

-- Lose control over governance

-- Start over

Artemis is Kore.ai’s answer to that problem.

This is not just another AI platform.

It is a ground-up rebuild focused on one idea:

AI should not just assist. It should build, govern, and optimize itself.

Prasanna shared something interesting during the conversation.

They didn’t evolve the platform.

They rebuilt it from scratch around what enterprise AI actually needs in 2026:

-- AI building AI

-- Built-in governance, not bolted on

-- Optimization as a continuous loop, not an afterthought

-- Designed for regulated industries from Day 1

And this is where it gets real.

Most enterprises today already have Amazon Web Services or Microsoft.

But the gap is not infrastructure.

The gap is:

How do you go from AI experiments to reliable, governed, production systems at scale?

That’s the layer Kore.ai is going after.

Also, one insight from Prasanna that stayed with me:

The biggest mistake is thinking AI is a model problem. It is actually a systems problem.

This launch is a signal.

We are moving from:

“Let’s try AI”

To:

“Let’s run the business on AI systems we can trust”

I’ll be dropping the full interview soon on The Ravit Show where we go deeper into:

-- Why they rebuilt everything

-- What “AI building AI” actually means

-- Where enterprise AI is headed in the next 18 months

This one is worth paying attention to.

#data #ai #koreai #agents #theravitshow

AI infrastructure conversations usually stay very technical. But my chat with , Santosh Erram, VP Partnerships DDN at Google Cloud Next '26 on The Ravit Show went in a different direction.

He kept bringing it back to one thing. Business value. Yes, compute is growing. Yes, GPUs are everywhere. But that is not the real bottleneck anymore. Data is. If you cannot move it fast, access it easily, and actually use it, your AI investment does not translate into outcomes.

What stood out was how fast things are moving. Their partnership with Google Cloud went from idea to launch in under six months. And now they are pushing things like 10 terabytes per second performance and hybrid tiering to meet real enterprise demands.

But the real proof was in the use cases.

- Salesforce pushing GPU utilization from around 48% to over 90%.

- Resemble AI driving cost savings.

- Sony Honda Mobility using it for autonomous driving.

Even financial firms bursting massive workloads into the cloud, hitting petabyte scale in a single day. This is not experimentation anymore. We are moving from AI pilots to real production. And the shift from training to inferencing is going to define the next phase. My biggest takeaway. AI is no longer limited by models. It is limited by how fast and how well you can work with your data.

#data #ai #ddn #infrastructure #googlecloudnext #api #google #theravitshow

AI is moving fast, but after my conversation with Alex Bouzari, Co-Founder and CEO at DDN, at Google Cloud Next '26, one thing became clear.

The bottleneck is no longer the model.

It is the infrastructure behind it. Alex broke it down in a very real way. Today’s AI systems are powerful, but the way data moves through them is still inefficient. You train these large models, but when it comes to actually running them at scale, things slow down. Latency increases, costs go up, and performance becomes unpredictable.

That is what is broken.

He shared how this shows up in real scenarios. When enterprises deploy AI, especially with large models, they struggle with speed and consistency. It is not that the model cannot perform, it is that the infrastructure cannot keep up with the demand.

At Next, DDN focused on solving exactly this. Building what Alex called a new foundation for AI, designed for high-performance workloads where data access and speed matter just as much as the model itself.

One concept that stood out was KV cache.

It sounds technical, but the idea is simple. Instead of recomputing everything every time a model runs, you reuse key pieces of information. That reduces latency and makes systems faster and more efficient. In large-scale AI systems, that becomes a big deal.

The bigger shift here is clear.

We are moving from experimenting with AI to operationalizing it at scale. And that means infrastructure is becoming the deciding factor.

What makes DDN different is their focus on this layer. Not just enabling AI, but making sure it actually performs in real-world environments.

My takeaway. The future of AI will not just be defined by better models. It will be defined by better infrastructure.

#data #ai #ddn #infrastructure #googlecloudnext #api #google #theravitshow

The man, the legend, CTO of Qlik, Sam Pierson. Always love chatting with him and this time I asked him some hard questions. I like the depth of this conversation. Thanks Sam for always being such a great sport and sharing some enterprise gaps in the Data & AI World :)#data #qlik #ai #qlikconnect #theravitshow

I had the chance to discuss key takeaways with Mike Capone, CEO of Qlik at Qlik Connect 2026 on The Ravit Show. What made this conversation stand out was how real it felt. No hype, no buzzwords. Just a clear view of where things are actually going. The shift is happening fast. We are moving from AI that gives answers to AI that takes action. And that sounds simple, but when you unpack it, it changes everything. It changes how data is prepared, how systems are designed, and how much trust you need before letting AI operate inside real workflows.

We talked about what is making this possible now, and why bringing analytics, data engineering, and trust together is no longer optional. It was also interesting to hear how companies like UPS, Schneider Electric, and HelloFresh are already moving from insights to execution.

One point that really stayed with me was this. Perfect models are not the goal. Impact is. And the teams that understand this are the ones moving faster.

We also spoke about trust, which is becoming the foundation for everything. Because once AI starts taking actions, you cannot afford to get it wrong.

And I ended with a simple question. What is Qlik’s role in the AI stack today. The answer was sharp and very telling..

#data #ai #qlikconnect #qlik #daredevil #api #trust #dataquality #agentic #agents #theravitshow

Spent time at Qlik Connect this week and one thing became very clear to me. Everyone is talking about AI, but very few are talking about what actually makes AI work. I had a great conversation with Sean Stauth and Kyle Jourdan from Qlik, on The Ravit Show and we went beyond the usual AI hype. What stood out to me is that most teams are not failing at AI because of models. They are getting stuck on data. Not because they don’t have data, but because they don’t trust it, can’t access it easily, or simply can’t operationalize it fast enough.

That gap between “we have data” and “we can actually use it for AI” is where most projects slow down. We also spoke about the constant tension between speed and foundations. Everyone wants to move fast with GenAI, but if your data layer is weak, you are just scaling confusion. The real challenge is not choosing between speed or building the right foundation. It is figuring out how to do both at the same time.

Another point that stayed with me was around agentic AI. Grounding LLMs in enterprise data is no longer optional. It is the difference between something that looks good in a demo and something that actually works in production. And again, it all comes back to data quality, governance, and accessibility.

My biggest takeaway from this conversation is simple. AI is no longer the hard part. Data is. The teams that figure this out will move ahead very quickly. The rest will keep experimenting without real impact.

Conversations like this are exactly why I enjoy being on the ground at events like Qlik Connect. Learn from them below!!!!

#data #qlik #ai #qlikconnect #theravitshow

AI is not the problem. The foundation is. That’s exactly what we unpacked in this conversation with Bruno BILLY and Rahul Bakhshi at APGAR On Air. Here’s what stood out for me. Most organizations are pushing hard on AI. But the data underneath is still fragmented, inconsistent, and not ready for scaleAnd that’s where things breakIn this conversation, we go deep on- What a real data foundation actually looks like- Why most AI initiatives fail before they even start- Where things break at scale across governance, ownership, and trust- What teams can realistically do in the next 90 daysIf you are building AI that needs to run in production, this is worth your timeWatch the full conversation here: https://lnkd.in/g-VSBqat#data #ai #apgar #theravitshow