The boom in artificial intelligence has made startups a minute. But as the dust settles, two once-hot business models look like cautionary tales: LLM wrappers and AI aggregators.
Darren Mowry, who leads Google’s global startup organization across Cloud, DeepMind, and Alphabet, says that starting with these hooks “lighted the test engine”.
LLM wrappers are essentially startups that wrap existing major language models, such as Claude, GPT, or Gemini, with a product or UX layer to solve a specific problem. An example would be a startup that uses AI to help students learn.
“If you’re really relying on the background model to do all the work and you’re almost white-washing that model, the industry doesn’t have a lot of patience for that anymore,” Mowry said in this week’s episode of Equity.
Wrapping up “a lot of smart stuff around Gemini or GPT-5” signals that you’re not isolating yourself, Mowry said.
“You have to have deep, wide channels that are horizontally separated or something that’s very specific to a specific market” for a startup to “develop and grow,” he said. Examples of deep LLM wrappers include Cursor, a GPT-enabled coding assistant, or Harvey AI, an official AI assistant.
Techcrunch event
Boston, MA
|
June 9, 2026
In other words, developers can no longer expect to slap a UI over GPT and get in touch with their product, as they could, perhaps, in the middle of 2024 when OpenAI launches its ChatGPT store. The challenge now is to create sustainable product value.
AI integrators are a subset of wrappers — they are startups that combine multiple LLMs into a single interface or API layer to query across models and give users access to multiple models. These companies typically provide an orchestration layer that includes a monitoring, governance, or measurement tool. Think: AI search startup Perplexity or developer platform OpenRouter, which provides access to multiple AI models through a single API.
While many of these platforms have gained ground, Mowry’s words are clear to entrants: “Don’t get into the merger business.”
In general, integrators don’t see much growth or progress these days because, he says, users want “some intellectual property built in” to make sure they’re being moved to the right model at the right time based on their needs — not because of behind-the-scenes computing or access barriers.
Mowry has been in the cloud game for decades, stints at AWS and Microsoft before setting up shop at Google Cloud, and he’s seen how this plays out. He said the situation today reflects the early days of cloud computing in the late 2000s/early 2010s as Amazon’s cloud business took off.
During that time, a crop of startups emerged to resell AWS infrastructure, marketing it as an easy entry point that provided tools, payment integration, and support. But when Amazon built its own business tools and customers learned to manage cloud services directly, many of those startups were squeezed. The only survivors are those who have added real services, such as security, migration, or DevOps consulting.
AI integrators today face the same margin pressure as model providers grow into business entities themselves, potentially sidelining middlemen.
For his part, Mowry is strong in vibe coding and developer platforms, which had a record year in 2025 with startups like Replit, Lovable, and Cursor (all Google Cloud customers, per Mowry) attracting significant investment and customer traction.
Mowry also expects strong growth in direct-to-consumer technology, from companies putting some of these powerful AI tools in the hands of customers. He highlighted the opportunity for film and TV students to use Google’s Veo AI video generator to bring stories to life.
Aside from AI, Mowry also thinks that biotech and climate technology are temporary — both in terms of business investment going into these two industries and the “amazing data” startups can access to create real value “in ways we couldn’t do before.”
