Monday night, Splunk kicked off its Welcome Keynote for attendees. It featured 2400 partners, customers and other guests, a keynote featuring Kamal Hathi, SVP and GM at Splunk and another highlighting Jeetu Patel, President and Chief Product Officer at Cisco. The focus was on outlining Splunk and Cisco’s product and AI vision and the latest tech trends shaping the future.
“This is my first time here at this incredible event, which is only possible because of this amazing community,” Hathi told the kickoff keynote. “That’s because of the versatility of the things all of you are building across these domains amid all these different industries. It’s the sheer scale of what you make possible.”
Hathi also gave attention to Splunk’s focus on AI.
“With AI, all of our approach is resilience,” he stated. “It’s about making AI trustworthy. If you can’t trust it, you can’t use it.”
Hathi also pointed out that the product keynote was early tomorrow at 830 am.
“There you can try out all the things they show on stage today,” he told his keynote audience.
Hathi was followed by Jeetu Patel, President and Chief Product Officer of Cisco.
“Last year we said – Cisco please don’t screw up Splunk,” he said. “It was a very low bar.”
Patel then went through the evolution of AI and where Splunk fit into it.
“We are in the next area of AI,” he said. “About three years ago, when Chat GPT came around, that was the first. The second era was about automational workflows, and the exponential capacity of throughput per capita.” The third era was about what Patel called the data gap.
“You should be thinking more about these three impediments,” Patel told his audience.
“First, there is an infrastructure constraint to meet demand,” he said. “Agents now work 7×24, which is more persistent demand compared to the past, and relies on deep research. The value, you will see, goes up.”
The second area, the workflows, is all about trust.
“These models introduce threat vectors, which have a mind of their own,” Patel said. “You need to build guardrails around them, because in their absence, most people won’t use these systems.”
The third area, the data gap, also has some problematic elements.
“On balance, the industry has done a pretty good job with the talent gap,” he stated. “On the other hand, it hasn’t done as good a job on machine data. We believe that the world needs to address the data gap around machines.”
The good news, Patel emphasized, is that Cisco is working on all three of these head on.
“We want to promise that Cisco will be the critical infrastructure for this era,” he said. “We need to be sure that workplaces get future proofed and it’s all digitally resilient. This challenge of digital resilience is a hard problem to solve.”
Patel noted that while the industry spends an awful lot of time on investigation and detection, with Cisco, they will be able to make remediation as fast as possible.
“The telemetry from Cisco is being worked into Splunk, as is the ability to ingest Cisco firewalls for free into Splunk,” he said, with the final point bringing on cheers from the audience.
“The future of the SOC will be agentic, as well as the ability to cure AI itself,” Patel added. “It does, however require the right level of visibility. It’s the tip of the iceberg to harness the full potential of AI. Splunk has always been a leader at unlocking. Now we want to do it for AI, to detect the undetectable, react autonomously and anticipate the future. It’s hard without good AI because of poor data quality and multiple data siloes.”
A key solution, Patel stated, is the announcement of Cisco Data Fabric.
“This is brand new, and at a revolutionary scale,’ he said. “We want to be able to unlock private data for AI and to unify the experience for humans and agents. This can be costly. However, if we bring Splunk to the data with the federation that we launched this last year, this can change. We now want to expand the breadth of the federation, and to this end, we have partnered with Snowflake. It will be the place to search for business-relevant data. You will be able to bring data together without moving it to a giant repository.” Azure and Cisco are two key examples of this.
Patel said that unlocking tends to be proprietary for AI.
“Most can’t create a moat to get the most value from their data,” he said. “We will teach AI to speak machine data – like it does natural language. We will be able to predict things we could never predict before as a result, and see and respond to events we couldn’t see before.”
Other news stemming from the event include the announcement of an open source time series model, from Hugging Face, which will be due in November. Splunk also plans to create a machine data lake.
“The machine data lake will be turnkey and optimized for AI learning,” Patel said. “You will be able to glean data in a way that was never possible before. In addition, thanks to AgenticOps, humans will also be able to collaborate with machines. This includes putting AI Canvas right into the Splunk interface.”
Patel concluded that the community has completely reimagined troubleshooting.
“You will be able to build your own GPT with AgenticOps,” he said. “Splunk is the machine data fabric for the AI era, and with the amount of information in the pipeline. we are just getting warmed up.”