AI Agents Replace Hedge Fund Staff With Tenfold Productivity

Hedge-fund manager Brian Kelly has his staff working around the clock, seven days a week, yet his annual payroll costs are a mere fraction of what they once were. The reason? Every member of his team is an artificial intelligence agent.

Kelly, who previously ran a cryptocurrency hedge fund, created his new trading firm Bracket22 to be powered entirely by agentic AI. The results, he says, have been nothing short of transformative for his bottom line.

The Cost Revolution: From $5 Million to $40,000

At his previous firm, Kelly employed seven to eight people across the globe, with many based in New York. Between salaries, compute costs, healthcare, and office space, his total labor-related expenses ran approximately $5 million per year.

Today, operating Bracket22 with AI agents, Kelly estimates his total annual costs at just $30,000 to $40,000 — a staggering reduction of over 99 percent.

“Now, when I’m using AI, I run somewhere around $30,000 to $40,000 a year, total,” Kelly told CNBC. “And that’s with every AI agent, that’s with all my compute, that’s with everything I need to completely replicate a hedge fund with AI.”

Meet the AI Workforce

Kelly has designed each AI agent to function as a specialist in a distinct domain, much like a traditional hedge fund structures its teams:

• Steffi — handles technical analysis, evaluating chart patterns and market indicators
• Desmond — manages quantitative strategies, running complex mathematical models
• Houston — serves as mission control, integrating insights from all agents and coordinating the final output

“I’ve crafted each of these agents to be a specialist in their field,” Kelly explained. “I wanted to isolate them and I wanted to get their unbiased view on what I’m doing.”

Crucially, Kelly retains human oversight in the decision-making process. The agents provide analysis and recommendations, but Kelly applies his own judgment to make the final call — a hybrid model that balances machine efficiency with human intuition.

Wall Street’s Broader AI Adoption

Bracket22 may be an extreme example, but it reflects a sweeping transformation across the financial industry. Major institutions are rapidly integrating AI into their operations:

• JPMorgan Chase CEO Jamie Dimon said in February that AI was already reshaping his workforce, with the bank planning to launch autonomous AI agents later this year capable of working independently for hours at a time. Dimon described “huge redeployment” plans for affected employees.
• Morgan Stanley is similarly routing significant portions of work to AI systems, streamlining operations across multiple divisions.
• Goldman Sachs has taken a more cautious stance, with at least one partner publicly warning about the dangers of allowing AI to erode bankers’ core reasoning and analytical skills.

Productivity Gains Beyond Cost Cutting

While the cost savings at Bracket22 are dramatic, Kelly emphasizes that the true opportunity lies in augmentation rather than pure replacement. He estimates he is “at least 10 times more productive” with his AI agents compared to his traditional team.

More importantly, he believes the same multiplier effect applies to existing workforces:

“If you take a staff of 100, with AI you’ve got a staff of a thousand,” Kelly said. “It’s not necessarily just, hey, you can replace everybody with AI agents. You can make your existing employees at least 10 times — maybe more — more productive.”

The Structural Shift in Finance

The implications extend far beyond a single hedge fund. The financial services industry has historically been labor-intensive, with significant overhead tied to analysts, traders, compliance officers, and support staff. AI agents capable of performing these roles continuously — without breaks, vacations, or burnout — fundamentally alter the economics of running a financial firm.

For smaller firms and solo operators, the Bracket22 model offers a blueprint for competing with institutions that have far greater resources. A single manager augmented by specialized AI agents can potentially match the analytical output of a much larger team, democratizing access to institutional-grade capabilities.

Risks and Considerations

Despite the enthusiasm, the shift toward AI-driven finance raises important questions:

• Skill atrophy — As Goldman Sachs has warned, over-reliance on AI could degrade the analytical reasoning skills that have long defined elite finance professionals
• Oversight gaps — AI agents operating autonomously for extended periods may make decisions that a human reviewer would catch, particularly in volatile market conditions
• Regulatory uncertainty — Financial regulators are still grappling with how to oversee algorithmic and AI-driven trading, and frameworks remain in development
• Concentration risk — If many firms adopt similar AI models and strategies, markets could see correlated behavior that amplifies volatility during stress events

What This Means for Business Leaders

The Bracket22 experiment offers several takeaways for business leaders across industries, not just finance:

Start with specialists, not generalists. Kelly’s approach of building focused AI agents for specific functions — rather than one AI trying to do everything — mirrors how effective human teams are structured. Specialization produces better results.

Maintain human checkpoints. Even in a firm powered entirely by AI agents, Kelly insists on human judgment for final decisions. This hybrid model preserves accountability and catches edge cases that automated systems might miss.

Reframe AI as augmentation first. While the cost savings are compelling, the productivity multiplier — making existing employees ten times more effective — may deliver greater long-term value than pure headcount reduction.

Reinvest savings into capability. The capital freed by reduced labor costs can be redirected into better AI infrastructure, data acquisition, and strategic initiatives that further compound competitive advantage.

The Road Ahead

Bracket22 trades only Kelly’s own capital and operates across cryptocurrencies, stocks, and commodities. While the firm is a solo operation today, the model it demonstrates could scale rapidly. As AI agents become more capable and affordable, the threshold for launching a sophisticated trading operation drops dramatically.

For the broader business community, the message is clear: the question is no longer whether AI will transform operations, but how quickly organizations can adapt to a landscape where a team of AI agents can match — or exceed — the output of a traditional workforce at a fraction of the cost.

Whether through full replacement, as Kelly has done, or through aggressive augmentation, the financial industry is leading a shift that will reverberate across every sector of the economy. Business leaders who understand and act on this transformation will be positioned to thrive; those who hesitate may find themselves competing against leaner, AI-powered rivals they never saw coming.

---

Edited by Palawan @QUE.COM
Website: https://QUE.COM Intelligence
Sponsored by: https://MAJ.COM AI Autonomous