The Ravit Show

Most AI tools are still just indexing documents. The Teamwork Graph has 150 billion connections. That difference is the whole game. I sat down with Jamil Valliani, the Head of AI Product at Atlassian, during Team '26 on The Ravit Show to understand why they are betting the next decade on this approach. Twenty years in Search before this role. Long before vector databases were trendy. Long before RAG was an acronym.A few things we got into:What the Teamwork Graph actually is. Why this architectural choice separates Atlassian's AI from everything else in the market.150 billion connections vs document indexing. Most enterprise AI tools search your text. This connects people, work, decisions, and outcomes across systems. The gap is bigger than I realized.Why connected data wins on accuracy. Atlassian's internal benchmark: 44% more accurate results using 48% fewer tokens. We broke down what is actually happening under the hood.A Search veteran's read on this moment. What makes this AI shift different from every other one. The most grounded take I heard at any conference this year.The line that stayed with me: in the next era of work, the company with the best context will win. Not the company with the best model. Models are getting commoditized. Context is not.If you work on retrieval, RAG, or graph-based AI inside an enterprise, this one is for you.#data #ai #atlassian #team26 #theravitshow

At SAS Innovate, I had the chance to speak with Reggie Townsend about one of the most important topics in AI right now.

Trust. What stood out from our conversation is how the market is evolving. Enterprises are moving quickly with AI, but the real challenge is no longer building models. It is ensuring visibility, transparency, and control as these systems start influencing real decisions.

We also discussed SAS’ new AI Navigator and why it matters. It is not just another layer. It is a way for organizations to understand where they are in their AI journey and how to move forward in a structured, governed way. This becomes critical, especially in industries where accountability is not optional.

My biggest takeaway.

If trust does not scale, AI will not scale.

Learn more from the interview below!!!!

#data #ai #SASInnovate #SASVisionary #theravitshow

From SAS Innovate, continuing conversations with leaders shaping how enterprise AI actually gets deployed. I spoke to Marinela Profi from SAS and this one cut straight to where the industry really is right now.

There’s a lot of excitement around AI. But most enterprises are still not ready to scale it. We talked about why. The gap is no longer about models. It’s about systems, governance, and how AI fits into real enterprise workflows.

One of the most interesting parts of the discussion was around MCP (Model Context Protocol) inside SAS Viya. This is about giving AI systems the right context, control, and structure so they can operate reliably in production environments.

Because without that, AI stays stuck in experimentation. We also went deep into why SAS is building dedicated agent infrastructure instead of just layering AI on top of existing tools. That decision matters.

It allows enterprises to move faster, while still maintaining control, auditability, and trust. That balance is what most organizations are struggling with today.

My biggest takeaway. The industry is moving from generative AI experiments

To governed, production-ready intelligence. And that shift requires a completely different approach to architecture.

#data #ai #SASInnovate #SASVisionary #theravitshow

Last week at SAS Innovate, I spoke with Dan Soceanu about something every enterprise is talking about, but very few have solved. AI-ready data.

The conversation was very practical. AI needs data, but not just more data. It needs data that is trusted, governed, and fit for purpose, especially as automation and agents start making decisions. We also went into digital sovereignty, which is becoming a key concern. Organizations are thinking deeply about where their data lives, how it is controlled, and how it aligns with regulations across regions.

What stood out is that this is not a future problem. It is a current one. And looking ahead, the focus is shifting from collecting data to making it usable and reliable for AI systems. My biggest takeaway.

AI success will depend more on data discipline than model sophistication.

#data #ai #SASInnovate #SASVisionary #theravitshow

I’m here at SAS Innovate, continuing conversations on what’s next for enterprise AI. I had a blast chatting with Amy Stout, and this one was focused on something many enterprises are curious about but not fully ready for yet. Quantum AI.

The discussion was very grounded.

We talked about the real barriers enterprises face today:

- Access to quantum systems

- Lack of expertise

- And uncertainty on where it actually fits in business problems

What SAS is doing with Quantum Lab is interesting because it is trying to remove that friction. Making quantum more accessible, more practical, and connected to real use cases.

The key takeaway for me was this. Quantum is not about replacing AI. It is about expanding what problems we can solve.

And while it may still be early, the groundwork being laid now will define who is ready when it scales.

That’s the kind of long-term thinking I’m seeing here at SAS Innovate.

More content coming from SAS Innovate on The Ravit Show.

#data #ai #SASInnovate #SASVisionary #theravitshow

I’m here at SAS Innovate, speaking with leaders who are shaping how enterprise AI actually gets deployed. Just spoke to Alyssa Farrell from SAS on The Ravit Show, focused on how SAS is accelerating AI across industries like financial services, public sector, and life sciences.What stood out was how clearly this is not about generic AI anymore. We talked about pre-packaged agents and industry-specific models, and why they matter. Most enterprises don’t struggle to build models. They struggle to make them work in real environments.Regulation, workflows, and domain complexity are not edge cases. They are the system.SAS is leaning into this by embedding that context directly into AI systems, which is what makes agentic AI actually usable at scale.This is the shift I’m seeing here. From building AI. To deploying AI that understands the businessStay tuned for more content on The Ravit Show.#data #ai #SASInnovate #SASVisionary #theravitshow

AI conversations are everywhere, but what stood out to me in my chat with Harmeen Mehta from Equinix at Google Cloud Next '26 was how grounded their approach is. They did not start with big external announcements. They started inside.

Harmeen shared a simple idea. If AI is going to change how a company works, it has to show up in how employees work first. Not as a side experiment, but as part of daily workflows. That shift is what moved AI from a pilot to something core to the business.

At Equinix, AI is not sitting on the edges. It is being used to remove real friction from day-to-day work. Helping teams move faster, reduce repetitive tasks, and focus on higher value problems. That is where the impact starts to become real.

But what stood out even more was how they approached trust.

Employee hesitation is real. Questions around accuracy, reliability, and job impact come up quickly. Instead of ignoring that, they leaned into it. Clear use cases, transparency, and gradual rollout made a big difference in adoption.

The biggest takeaway from this conversation was simple.

Do not try to scale AI before you make it work internally.

If your own teams are not using it, trusting it, and seeing value from it, scaling it across the business will not work.

And looking ahead, the shift is already happening. Not years from now, but right now. AI is starting to change how work gets done inside enterprises, one workflow at a time.

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

Just wrapped a great conversation with Woon Ho Jung, CTO - Cloud Native, Commvault, at Google Cloud Next 2026 and this one hit a nerve. Everyone is talking about multi-cloud, AI pipelines, and scaling data.

But almost no one is talking about what’s quietly breaking underneath it all. Data protection. We got into what’s really happening inside enterprises today.

Teams assume replication and retention policies are enough. They’re not.

At scale, across billions of objects, things get messy fast. Gaps show up where you least expect them.

That’s where the big announcement comes in. Clumio is going deeper with Google Cloud. Clumio for GCP is not just another backup solution. It’s a rethink of how you protect cloud-native data, especially inside Google Cloud Storage where most AI and analytics pipelines live today.

What stood out to me:

- Protecting data at massive scale is still an unsolved problem for many teams

Native tools give a false sense of security

- Resilience in the AI era needs a completely different approach

If you’re building on Google Cloud right now, this is something you need to pay attention to. This is not about backup. This is about trust in your data layer.

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

AI sounds exciting… until you actually try to use it inside a company. That was my biggest takeaway from my conversation with Michael Fasulo from Commvault at Google Cloud Next '26 on The Ravit Show.

Everyone wants AI, but when it comes to real deployment, things break. Data is messy, systems are disconnected, and trust is missing. The gap is not ambition, it is readiness.

One line that stayed with me. If your data is compromised, your AI is compromised.

And with agentic AI, it gets even more real. These systems are not just answering anymore, they are taking actions. That means mistakes can have real impact.

My takeaway is simple.

The companies that win will not be the ones trying the most AI. They will be the ones fixing their data and putting the right guardrails in place first.

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

Test