There Was HerculesAI

Since 2017, HerculesAI (previously known as Zero Systems) has been focused on automating professional services, initially targeting the legal sector. As part of this endeavor, the company has been developing large language models for several years, well before the concept became widely recognized. This early start positioned them advantageously when ChatGPT burst onto the scene in late 2022, sparking widespread interest in LLMs.

The company has recently secured a $26 million Series B investment to capitalize on its current momentum and further expand its operations.

Alex Babin, the CEO and co-founder, reveals that they had been working with smaller models, ranging from 500 million to 2 billion parameters, since around 2020. These models were designed to run on edge devices for compliance reasons. However, Babin notes that before ChatGPT’s emergence, this aspect of their solution received little attention from others.

HerculesAI’s prior experience with LLMs proved advantageous when interest in the technology surged following ChatGPT’s debut. Babin describes the shift in market perception, “It was maybe eight or nine months before ChatGPT, and I remember speaking to our clients, explaining to CIOs what an LLM is — and no one cared,” Babin said.

By November of that year, this attitude had dramatically changed, with widespread interest in the concept driving rapid business growth over the subsequent year.

Currently, the company employs several models for three primary functions: data extraction, transformation, and verification. While data extraction is a common feature, and data transformation involves automatically creating rules and structures, Babin emphasizes the particular importance of the verification process. He said, “It’s really the holy grail when you can compare information extracted and then transform it to the source of truth, whether that’s regulations, policies, contracts, laws, or anything,” 

This verification ensures that any discrepancies with source materials are automatically flagged for employees.

These three components have enabled the startup to develop a multi-agent system that automates these activities. Babin explained, “Those multi-agent systems can be applied to high-value, continuous processes or workflows that require “automated” decision-making.” 

These capabilities are crucial for HerculesAI’s primary regulated industry clients. Their customer base now extends beyond legal services to include insurance and financial sectors.

The company’s AI approach seems effective, with reported quadruple growth in the past year. Their clientele includes 30% of the top 100 U.S. law firms, as well as numerous Fortune 500 companies such as Mercer, Standard & Poor’s, and State Farm.

With approximately 75 employees, Babin plans to maintain a lean operation despite the new funding. He intends to focus on enhancing internal processes rather than expanding the workforce:

“I don’t see why we need to hire more people. We will actually invest more in our internal processes and automation. We have to eat our own dog food and use our own products to make ourselves more scalable,” he said.

The recent funding round was led by Streamlined Ventures, with contributions from Proof VC, Thomson Reuters Ventures, Alumni Ventures, and various industry angels.