Cisco, Trace3 Talk AI Readiness as AI Budgets Expand

Channel partners are finding their lanes for helping business customers with artificial intelligence (AI) plans and platforms.

Three months into 2024, AI is still the talk of the town in enterprise IT. Vendors are rolling out AI-flavored solutions regardless of their place in tech stack. And most importantly, customers are demanding it.

"Whether it's the board level or the C-suite that's putting pressure on the organization to 'do something with AI,' it's causing this frenzy of activity," Trace3 chief technology officer Tony Olzak said. "We're seeing the budgets open up; we're seeing the ideas go crazy. Imaginations are running wild."

Trace3's Tony Olzak

Jason Gallo, Cisco's vice president of global partner and routes to market sales, said demand is so high that customers are actually expanding technology budget, citing IDC research. And partners are salivating at the opportunity.

"That's something that we haven't always seen. Normally there is an expectation to transfer one part of the budget to another to the needs of the company," Gallo told Channel Futures. "But in the current state that we're in, you have to add more budget if you want to move as fast as you need to. And I think that's a great opportunity for the channel partners."

Cisco's Jason Gallo

But in many cases, customers don't exactly know what they're demanding. While Cisco's recently revealed AI Readiness Index found that 95% of businesses have built or are building an AI strategy, only 14% said they are ready to actually onboard AI.

"'Do something new with AI' doesn't really give you any guidance on what you should do," Olzak said. "And as all these ideas are coming, how do you help clients through their strategy?"

Talking AI Readiness

Olzak and Gallo were part of a panel at the 2024 Channel Partners Conference & Expo to walk partners through frameworks they should consider before engaging in AI. They focused on three pillars: infrastructure, data and responsible AI.

Networking vendors like Cisco, HPE and Juniper say large language models (LLMs) and the data they require makes their infrastructure more important than ever. Gallo said previously, network teams would question if their use cases really required them to 400-800 Gigabit capacity. Now businesses are moving en masse to those features, Gallo said.

But Cisco doesn't want to pigeonhole itself on infrastructure when it comes to AI, Gallo said. One solution underpinning that ambition is the Open AI Dashboard the company is developing within its Full Stack Observability framework. The idea is to track usage of LLMs within a company. Gallo said Cisco wants to address a new kind of shadow IT. The rise of cloud-based SaaS applications led to purchases by employees that often flew under the radar of the official IT purchasers.

"Customers were saying, 'I'm spending way too much, and I didn't even know it because everyone is spinning it up. Cloud is so accessible.' We think the same thing it's going to happen with AI," Gallo said.

Cisco's Outshift incubation business recently unveiled Motific, software designed to give more visibility into generative AI deployments.

And for Trace3, the channel partner isn't stopping at infrastructure either. Olzak said the company had already built resources to consult closely with end customers at the data level.

"We've been in the data space for years and have a large data engineering team. We do everything from data strategy to data governance, to building modern data platforms to creating new data pipelines, and we've also been in the security space for a long time," Olzak said. "These are interesting building blocks where, if you were already ready, you don't have to get ready."

Responsible AI has grown in popularity in the last years, with committees deliberating over ethical concerns.

Gallo pointed to the area of machine learning engine training. He said Cisco has created more than 20 million synthetic emails that its AI threat intelligence can use to train itself, rather than using existing emails that may have come from customers.

Skills Evolution

Cisco's Alex Pujols

Alex Pujols, vice president of global partner engineering at Cisco, said generative AI is prompting a shift in the skills customers, partners and vendors need from their engineers.

"I think being a true-blue networking engineer over the next 15 years isn't going to look like what it looked like in the last 15 years," Pujols told Channel Futures.

Pujols said he sees network engineers no longer focusing on Layer 2 and 3 switching and security.

"The stack that network engineers have operated with for almost 25 years is different than the AI and ML stacks we're seeing today. So you're seeing this interesting collision of traditional network engineers colliding with developers to create some new interesting hybrid type of roles, and they are extremely in demand right now," Pujols said.

Might there be a new name for this type of engineer? Pujols said the term "AIOps" is gaining some traction. He mused that a "cascade of additional roles" might occur in the way DevOps, NetOps, SecOps and others have popped up.

"Honestly I think it's still kind of finding its footing. Some people will say it's a network engineer with AI/ML skills. Sometimes we'll say it's a machine learning engineer with networking skills. We're so early that I don't know if it's been formalized to that degree, but it's happening," Pujols said.

Cisco for its part is adding specializations and training that will help engineers expand their skill sets. Moreover, the company is offering incentives to Cisco internal engineers around new AI and ML use cases that the company can support.

Advice for Partners

Pujols said partners will sometimes ask him to name three things they need to do to make the most of the AI wave.

But that's not so easy to answer.

"It's an impossible question to answer. It would probably be almost irresponsible for me to answer, because there's so much about AI that is really a consultative selling motion. It involves a combination of what that partner is, what the partner wants to be and what their customers need them to be," Pujols said.

Some partners may discover that they don't need to build customized LLMs and optimization algorithms. They may instead sell AI as part of a product.

"If you can isolate those things that synergize with your business and your customer demand, then in essence you put that mark on the map, and then that's when we can really help you," he said.

Olzak said the increased intrigue around AI (and budget) gives opportunity for all kinds of partners, because partners will play a wide range of roles.

"This is a great opportunity to actually provide true guidance as a partner versus selling whatever you happened to be selling," he said. "Because there are so many hats that have to be worn to turn this into value. There's an opportunity for you to get engaged somewhere along that journey."