As AI agents become a bigger topic in the cryptocurrency space, Pranav Ramesh told CoinDesk that Nasdaq has already been using them in various sections of its business and has greatly expanded that use over the past 18 months or so.
Ramesh, head of options research at Nasdaq and co-founder and CTO of Leadpoet, said the most significant change has been in sentiment. “AI agents are relatively new and have probably been used more and more in the last six months,” he said, arguing that previous systems too often failed for sensitive business workflows.
He said Nasdaq is using AI agents in areas including market surveillance, compliance and market microstructure analysis, and pointed to Nasdaq Verafin’s “Gentic AI Task Force,” which Nasdaq said automates “high-volume, low-value compliance processes” in anti-money laundering work.
Ramesh also pointed out Nasdaq’s AI-powered order type. Nasdaq announced in 2023 that its Dynamic M-ELO order type had become the first AI-powered swap order type approved by the SEC, using an AI model with more than 140 factors to adjust to market conditions in real time.
For Ramesh, that experience informs how he views cryptocurrencies. He said cryptocurrency trading platforms are likely to move aggressively with AI agents for both internal trading and retail tools, including position analysis, trade suggestions and execution support. “The cryptocurrency trading world is actually going to lead the fight over how AI is used in the retail environment,” he said.
He did not describe that change as completely autonomous. Instead, he said the model he sees taking hold is one in which agents handle most of the analysis and workflow, while humans retain final approval. In the interview, he said that at Nasdaq, many systems have not yet reached full automation, and human review remains the last step.
AI and AI agents will replace much of human work
Ramesh’s views are also unusually forceful on labor matters. “Yes, there will be a lot of jobs needed,” he said of AI agents, adding that he believes lower-level software, customer service and analyst roles are already being displaced as systems become faster, cheaper and more reliable. He raised it as an observable trend rather than a prediction.
And they seem to be right as companies, including most recently Crypto.com, which laid off 12% of its staff in a bid to achieve greater automation and efficiency through AI. Previously, cryptocurrency research firm Messari parted ways with several members of its staff and its CEO as the company transitioned to what the new CEO called an “AI-first company.” Last month, Block, the payments company founded by Jack Dorsey, announced plans to cut 40% of its employees, or more than 4,000 people, citing improved AI models.
The AI trend led to the founding of Leadpoet
That thesis also marked his path to Leadpoet, the startup he co-founded with Gavin Zaentz. According to a company fact sheet from February 2026, the two met at Nasdaq and founded the company after repeatedly encountering the same problem: Exit tools could generate static lists, but identifying actual purchase intent still required manual research.
Leadpoet describes itself as an AI-powered lead scoring platform that turns web signals and company context into “decision-ready lead recommendations,” emphasizing “accuracy over volume.” The company says it supports private deployments so customers can score intent and drive reach on their own data without exposing it to a vendor.
The fact sheet says Leadpoet uses Bittensor, which describes itself as a blockchain-powered decentralized AI network that allows participants to contribute models and compute while earning rewards. Ramesh said a competitive, decentralized structure is part of the appeal, because it can improve models faster than a centralized roadmap.
Leadpoet also says he is a member of NVIDIA Inception, NVIDIA’s startup program for AI companies. NVIDIA describes Inception as a free program that offers technical resources, marketing support, and access to its broader ecosystem.
In the company’s February 2026 fact sheet, Leadpoet says it hit an annualized run rate of $1 million in its first quarter after launch and was backed by DSV Fund and Astrid. In that same material, DSV Fund CIO Siam Kidd said Ramesh and Zaentz combine “deep AI engineering experience with a real understanding of day-to-day sales.”
Ramesh linked the company directly to what he says he’s seen within large institutions adopting AI: agents moving from assistants to systems that can handle real operational work. In the case of cryptocurrencies, he said, that change is likely to become visible faster than in many other corners of finance.
