#enterprise-ai

#enterprise-ai

Over the past few years, I've reviewed thousands of APIs across startups, enterprises and global platforms. Almost all shipped OpenAPI documents. On paper, they should be well-defined and interoperable. In practice, most fail when consumed predictably by AI systems. They were designed for human readers, not machines that need to reason, plan and safely execute actions. When APIs are ambiguous, inconsistent or structurally unreliable, AI systems struggle or fail outright.

Every year, TechCrunch's Startup Battlefield pitch contest draws thousands of applicants. We whittle those applications down to the top 200 contenders, and of them, the top 20 compete on the big stage to become the winner, taking home the Startup Battlefield Cup and a cash prize of $100,000. But the remaining 180 startups all blew us away as well in their respective categories and compete in their own pitch competition.

Jain said he had tried to automate internal workflows at Glean, including an effort to use AI to automatically identify employees' top priorities for the week and document them for leadership. "It has all the context inside the company to make it happen," said Jain, adding that he thought AI would "magically" do the work. The idea seemed simple, but it hasn't worked.

For many, enterprise AI adoption depends on the availability of high-quality open-weights models. Exposing sensitive customer data or hard-fought intellectual property to APIs so you can use closed models like ChatGPT is a non-starter. Outside of Chinese AI labs, the few open-weights models available today don't compare favorably to the proprietary models from the likes of OpenAI or Anthropic. This isn't just a problem for enterprise adoption; it's a roadblock to Nvidia's agentic AI vision that the GPU giant is keen to clear.

It is becoming increasingly difficult to separate the signal from the noise in the world of artificial intelligence. Every day brings a new benchmark, a new "state-of-the-art" model, or a new claim that yesterday's architecture is obsolete. For developers tasked with building their first AI application, particularly within a larger enterprise, the sheer volume of announcements creates a paralysis of choice.

They know all too well how Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla have delivered more than half of the S&P 500's gains in recent years, setting a high bar for everyone else to clear. But things change: One minute, Alphabet is behind the curve on AI and then Google's latest Gemini launch sparked a 'Code Red' from ChatGPT's Sam Altman.

"What's driving all of this is the awareness from CEOs and executives that this is the time to invest in AI," Jain said in an exclusive interview before the conference. "Everybody has been looking for a safe, secure, more appropriate version of ChatGPT for their employees. And we bring the capabilities that ChatGPT brings to consumers to business users, and in the context of their company."

OpenAI released new data Monday showing enterprise usage of its AI tools has surged dramatically over the past year, with ChatGPT message volume growing 8x since November 2024 and workers reporting they're saving up to an hour daily. The findings arrive a week after CEO Sam Altman sent an internal "code red" memo about the competitive threat of Google.

The launch comes as Mistral, which develops open-weight language models and a Europe-focused AI chatbot Le Chat, has appeared to be playing catch up with some of Silicon Valley's closed source frontier models. The two-year-old startup, founded by former DeepMind and Meta researchers, has raised roughly $2.7 billion to date at a $13.7 billion valuation - peanuts compared to the numbers competitors like OpenAI ($57 billion raised at a $500 billion valuation) and Anthropic ($45 billion raised at a $350 billion valuation) are pulling.

In this comprehensive interview at SAP Connect in Las Vegas, Stephan de Barse, President of Global Business Suite at SAP, makes the case for why best-of-breed enterprise software is being replaced by best-of-suite approaches, driven by the commoditization of the application layer through AI. De Barse, who previously spent seven years at o9 Solutions, a best-of-breed supply chain vendor, brings a unique perspective to the debate.

Enterprises can move from small pilots to full deployments without violating their jurisdiction's rules on where data should live. The reality is that, earlier, most security and compliance teams weren't rejecting GenAI because of model design; they were rejecting it because storing data in the US or EU pushed them into conflict with GDPR, India's incoming DPDPA norms, UAE's federal rules, or sector-specific mandates like PCI-DSS,

"You have to pay them a lot because there's not a lot of these people for the world," Gomez said. "And so there's tons of demand for these people, but there's not enough of those people to do the work the world needs. And it turns out that these models are best at the types of things those people do."

Everybody would say yes, but we kept moving the goalposts," Ghodsi said in the discussion, which was published Tuesday.

"The customer can set up their own tagging system or we can dynamically create that for them,"

Without really knowing how work is done, it is really hard to know where to improve it, where to automate it, where agents can help.

PromptQL is an enterprise platform that aims to automate some of the work of a typical consultant, like surfacing insights and generating reports. It helps clients build custom AI analysts by integrating their internal data with the foundation models they already use. Once deployed, these AI analysts can perform tasks typically handled by data scientists or engineers - and continuously learn and adapt to their environments over time.