A stack of fund documents that once cost $200,000 and months of lawyer time can now be drafted in minutes. Kingsley Advani, CEO of Allocations and Advani Family Office, recently joined TheStr
A stack of fund documents that once cost $200,000 and months of lawyer time can now be drafted in minutes.
Kingsley Advani, CEO of Allocations and Advani Family Office, recently joined TheStreet Roundtable during BVI Finance’s Fintech on the Seas to share why he has given every single one of his employees a monthly AI quota.
"AI has reduced our cost by 50%. And it's doubled our revenue," he said.
AI isn’t just a tool some of his teams use, it has become, in many ways, the foundation of his fund’s infrastructure.
"Every team member is plugged into AI," he said. "Every team member has AI credits that they need to spend every month. Sometimes they have multiple models — Claude, Grok, some of the Chinese models, GLM, OpenAI. We pick the best model for the purpose."
Those purposes are quite varying. Advani said that “95%” of his fund’s code is written initially with AI, saving massive amounts of money that would have been spent on software engineers.
The clearest savings show up in the paperwork. Compliance draft reviews that “used to take weeks” now run “in a few minutes for high-level reviews.” Legal drafts have compressed the same way.
"Let's say we have to draft BVI documents. Historically you may have had to spend weeks or months with a lawyer — maybe cost you $200,000," he said. "Now you can make a pretty good draft and get a legal opinion on it, and let them focus on reviewing it rather than drafting everything line by line," Advani explained.
The BVI example isn’t hypothetical. Advani set up his first entity in the territory a decade ago and now runs about 10 there, with plans to expand as clients demand SPVs and funds.
"The regulation has gotten a lot better here. The formations have gotten a lot faster," he said in an earlier conversation at the event.
Fraud monitoring is the same story. Advani even believes that LLMs can be a good place to share investment thesis’.
LLM’s, or large language models, are AI systems trained on massive amounts of text to predict and generate language. It is the tech that underlies all major AI platforms today.
"I ask it: is Bitcoin undervalued or overvalued? And it definitely presents some good arguments."
Behind the adoption push is a conviction about where the cost curve is going. Advani shared that the cost of intelligence had collapse “280 fold” in recent years. Stanford’s AI Index found that inference costs for GPT 3.5-level performance fell more than 280 fold between November 2022 and October 2024.
Investors are debating whether or not there is an AI bubble. Advani thinks there are good arguments on both sides, and refuses to pick one.
"I think they're undervalued and overvalued," he said. "In the short term there is some exuberance, but it is both a shift and a shock."
The shift drives his long-term view of the AI buildout.
"AI is transforming how the world operates, and every company will be plugged into AI — or every competitive company will be plugged into AI," he said. "Every person will be able to leverage AI for medical, for investment, for legal, for whatever you want. This will profoundly change how people operate."
The shock is what happens to share prices along the way.
"In any shift, there will be a bit of a shock as well," he said. "It's important to still look at the fundamentals of the assets — look at their revenue, their growth — and make sure that you have some margin of safety."
The AI buildout has not slowed down one bit in 2026. On Monday, August 10, 2026, Nvidiaannounced that it was partnering with the largest asset managers in the world including BlackRock, Apollo, Goldman Sachs, Blackstone, and more to mobilize more than $500 billion to help finance AI infrastructure.
