The enterprise AI hype cycle is over. Now it’s time for a reality check.

Earlier this month, we attended HumanX in Las Vegas and while the conference made it clear that AI technologies are advancing at breakneck speed, their integration into enterprises remains inconsistent. Leaders across industries recognize AI’s potential, but they face significant challenges in adoption, trust, and ROI measurement. The landscape is shifting: businesses that once chased AI for the sake of innovation are now demanding clear, measurable value.

Our experience at HumanX reaffirmed a key insight: the AI gold rush is not about who has the biggest model or the most GPUs. Instead, the real winners will be those who solve fundamental enterprise inefficiencies by embedding AI into core business processes.

So where should AI startups focus to maximize impact and investment potential? Let’s explore the key takeaways from the event.

Our AI Takeaways from HumanX

The conversations at HumanX made one thing clear: AI is no longer just a promising experiment—it’s a business imperative. Yet, while many enterprises are eager to adopt AI, most struggle to move beyond pilot programs and into full-scale deployment. 

Investors and AI leaders emphasized that success will come to those who embed AI deeply into enterprise workflows, leverage proprietary data, and deliver measurable ROI. Here’s what we took away from the event.

1. AI is an Enterprise Power Tool (Not Just a Chatbot)

One of the biggest misconceptions about enterprise AI is that it’s just an interface upgrade. Companies flooded the market with chat-based AI assistants, hoping to replicate ChatGPT’s consumer success. But as Mike Krieger, CPO of Anthropic, pointed out at HumanX, enterprises don’t need just another chatbot—they need AI that fundamentally improves how work gets done.

“We all keep saying, gosh, if it’s just chat boxes and chatbots, a year from now, we’ll all have failed. But then what?” ~ Mike Krieger, CPO of Anthropic

Krieger highlighted Claude Code, Anthropic’s AI-powered development tool, which gained 100,000 users in a week by deeply integrating into engineers’ workflows rather than forcing them into a new interface. This underscores a critical investment thesis: the most successful AI startups will be those that seamlessly fit into existing enterprise systems, rather than demanding companies overhaul their workflows.

For startups, this means focusing on vertical-specific AI applications—AI that enhances specific functions like sales forecasting, risk analysis, or customer support automation. The real opportunities lie in AI that acts as a power tool, not just a chatbot, allowing enterprises to scale their expertise rather than replace it.

2. From Thin Wrappers to Full-Stack AI Applications

The narrative around “thin AI wrappers” is changing. At HumanX, Steve Jang of Kindred Ventures emphasized that AI’s true value isn’t just in building wrappers around foundation models but in developing robust applications that integrate deeply into enterprise processes:

“You can swap out most of the LLMs today and have a very similar experience that the customer probably will not even discern… The magic is actually in that application layer.” ~ Steve Jang, Kindred Ventures

This shift means that startups need to go beyond generic AI-powered features and focus on solving mission-critical enterprise problems. Instead of building another general-purpose AI assistant, they should be asking:

• How does this AI tool help companies generate revenue or cut costs in a measurable way?

• Does it integrate seamlessly with enterprise software stacks like SAP, Salesforce, or AWS?

• Can it provide defensible value that goes beyond what OpenAI or Anthropic’s APIs already offer?

The most promising startups today aren’t just leveraging large language models (LLMs)—they are building entire AI-powered systems that tackle vertical-specific enterprise challenges in industries like healthcare, finance, and logistics.

3. The Future of Enterprise AI: Proprietary Data Is the Moat

One of the most compelling discussions at HumanX was around how startups can compete with AI giants like OpenAI and AWS. According to Josh Constantine, Venture Partner at SignalFire and technology journalist, the key differentiator is data:

“You can build crazy, incredible companies built on the top of these models, as long as you have proprietary data sources, a great human-in-the-loop feedback loop, and incredible speed to be able to build faster than these incumbents." ~ Josh Constantine, Venture Partner at SignalFire

Why does this matter? LLMs are becoming commoditized. While today’s AI race is often framed around model size and performance, the long-term value will come from who owns the data that makes these models truly useful in enterprise settings.

For AI startups, this means:

• Building exclusive access to high-value datasets in industries like legal, biotech, or financial services.

• Creating feedback loops where AI improves continuously based on real-world enterprise interactions.

• Developing defensible AI models that perform significantly better than off-the-shelf solutions in specific domains.

A great example of this approach is Perplexity AI, which took Meta’s open-source Llama model, fine-tuned it with proprietary data and launched its own branded AI assistant, Sonar. This kind of innovation—leveraging public models but enhancing them with proprietary insights—is where we see massive potential for AI startups.

4. The Real Money in AI: Workflow Automation, Not Just Chat Interfaces

A recurring theme was that AI’s biggest enterprise impact won’t come from replacing human workers but from supercharging their efficiency. As Nicole Zhang, Chief Business Officer at A-Team, describes:

“We helped a global healthcare provider develop an AI-powered document assistant that reduced nurse burnout by automating paperwork. Specifically, we saw a 40% boost in nurse productivity in just four months, freeing them to focus on vital patient care. The digital revolution gave us better interfaces. The AI revolution will give us better outcomes, but only for those ready to assemble the right talent, technology, and approach.” ~ Nicole Zhang, Chief Business Officer at A-Team

Startups that reduce friction in enterprise workflows—whether by automating data entry, streamlining compliance, or enhancing cybersecurity—are seeing the fastest adoption. For example, one AI-driven document assistant in healthcare increased nurse productivity by 40% in just four months.

The best enterprise AI startups are not just selling software—they’re selling business transformation. Investors are looking for companies that don’t just provide point solutions but offer end-to-end automation that directly ties to business outcomes.

5. AI Startups Must Prove ROI—Not Just Hype

The final and perhaps most critical insight from HumanX was the shifting expectations of AI investment. Investors no longer throw money at “AI-powered” companies without clear business models. Jai Das, of Sapphire Ventures, summed it up well:

“And the beauty of AI is actually you can build these companies much, much more efficiently. Having limited resources always helps you build better companies.” ~ Jai Das, Co-Founder of Sapphire Ventures

This marks a turning point in AI investing:

• Startups that can show clear cost savings or revenue growth will have a huge advantage.

• Investors are looking for capital-efficient AI businesses, not ones that burn billions chasing scale.

• Companies that focus on deep enterprise integration, automation, and data-driven AI will dominate the market.

For AI startups, this is the moment to prove value—not just innovation.

The AI Gold Rush is About Business Transformation, Not Just Technology

The AI revolution is not about who has the biggest model. It’s about who can turn AI into real enterprise value. As the HumanX discussions made clear, the winning AI startups will be the ones that:

✅ Embed AI deeply into enterprise workflows

✅ Build full-stack applications, not just thin wrappers

✅ Leverage proprietary data as a competitive moat

✅ Automate high-value business processes

✅ Show clear, measurable ROI

At Maxitech, we are investing in AI startups that transform industries—not just those that ride the AI wave. The future of enterprise AI isn’t just about chatbots—it’s about reshaping the way businesses operate at their core.

The gold rush is here. The real question is: are you building where it matters?