This Startup Is Bringing Ai Agents To Banks And Money Managers Including Ubs, Blue Owl Capital, And T. Rowe Price. Here's The Deck It Used To Raise $8 Million.

Chandini Jain, CEO of Auquan.

Auquan

• Auquan was founded in 2018 by Chandini Jain, a former analyst and derivatives trader.
• It launched an AI product in late 2023 that can automate research work usually done by analysts.
• It has since inked deals with UBS, Blue Owl Capital, and T. Rowe Price.

The wave of fintechs building autonomous agents for finance firms designed to do the work of a junior analyst or banker is swelling.

One such startup, Auquan, uses generative AI to automate the time-consuming but ubiquitous task of gathering and processing data and putting that information into a written template, like a due diligence report, an investment committee memo, or a pitch book.

Demand for this kind of technology is heating up in the finance industry, which is often bogged down with manual processes around data management, processing, and analysis. Since launching in October 2023, Auquan has brought in close to $2 million in annual recurring revenue and secured UBS, T. Rowe Price, and Blue Owl Capital as customers, according to CEO and cofounder, Chandini Jain.

From banking to software engineering to research, finance firms are keen to implement AI assistants that can carry out multi-step processes. While it's still too early to tell just how much the technology will impact adopters' bottom lines, venture-capital investors are writing checks to get in on the ground floor of what some industry leaders are calling a revolutionary technology.

In addition to scoring big-name clients and its revenue stream last year, Auquan also closed $8 million in seed funding. The round was led by Peak XV and included Neotribe Ventures.

Auquan has made inroads with various divisions at financial firms, from private-market investing to investment banking, as well as risk and compliance, and investor relations, Jain said. Across those functions, it's most heavily used by analysts or associates to produce documents or templates for their MDs, partners, or division heads.

Due-diligence reports are a big use case for Auquan. The startup automates the creation of 3,300 due diligence reports for 20 different clients, saving them a cumulative 55,000 hours of work, according to customer estimates.

How Auquan works behind the scenes

Auquan is built to try to mimic the humans whose jobs it is doing, Jain said.

The first step is accessing the raw data. Auquan pulls data from providers, like FactSet, CapIQ, and Pitchbook, as well as public data sets from government agencies and news sites. It can also plug into a client's internal file systems.

The second part involves the user stating an intent with an example, such as "I want to create an investment committee memo and I want it to come out looking like this template document," Jain said. Under the hood, the tech relies on an "agent super orchestrator" that breaks down the specific jobs to be done and organizes several "mini agents" to take on each of those jobs, Jain said.

In the investment committee memo example, there might be an agent that identifies the fields that need to be filled, another to run searches on underlying vendor data, another that scans public data, and a writing agent that takes all of the info and puts it into a company-specific format based on the template, like if a section should be presented in bullet points or a table. It's exported in the desired interface, such as a PowerPoint presentation or a Google Doc with the proper corporate branding. All of this happens automatically without human intervention, Jain said.

The first draft is presented to the user as a starting point. The user can make edits and tweaks for future documents, she said. The agent super orchestrator will assign new mini agents as needed, she added.

Pricing for Auquan is based on clients' estimated desired outcomes, Jain said. Examples of outcomes are producing one slide deck, one report, or one compliance check. Once the client chooses what workflow to automate, Auquan charges a dollar amount for that outcome, and multiplies it by how many times that process is expected to run, she said.

Too much data, not enough people

Jain knows firsthand how labor intensive it is to extract insights from data. Before Auquan, she worked as an analyst at Deutsche Bank and derivatives trader at the Dutch market-maker and proprietary trading firm Optiver, where she was drowning in information with not enough time or help to distill it.

"If I or anyone on my team could make the case for why we thought any data set would help us make better decisions, we could buy it no questions asked," Jain said. "What we didn't have a lot of was resources or time to go through that information," she said.

She would learn from conversations with financial clients that she wasn't alone in that problem. The broad applicability has won over investors.

Here's the pitch deck Auquan used to raise $8 million.