Herculean AI, a generative AI company innovating around the longer term of labor, has developed a brand new methodology for quickly deploying virtual AI employees within the enterprise. Additionally, the corporate formerly referred to as Zero Systems has introduced RosettaStoneLLM, a model that helps firms in regulated industries automate complex workflows that require cognitive decision-making.
In a so-called “assembly line process,” firms can select pre-built components to develop and deploy virtual AI employees. Everything is pre-built, tested and pre-configured upfront – no custom-made products. According to Hercules AI, this may create AI agents which can be “top quality, cost-effective and simply scalable.” All that may must be done could be to fine-tune the model in order that the bot follows the required workflow.
Most businesses may prefer a bot that’s tailored to their needs and specifications. However, in a regulated area this costs money and time, which might harm customers. However, those in finance, insurance and legal services could also be more acceptable as each component has been reviewed by a regulator to comply with laws and ensure data security. If it really works for others within the room, why not replicate it? The only adaptation allowed could be the Large Language Model (LLM).
Another offering from Hercules AI is RosettaStoneLLM. It is predicated on Mistral-7B and WizardCoder-13B and has 7 billion parameters. Companies using this may convert structured data from spreadsheets in order that it may be mapped and transformed by AI. Imagine the quantity of database exports and spreadsheets a regulated company could have. Converting these files into useful system-wide data will be costly and time-consuming. This LLM is designed to remodel large amounts of structured data into something consistent together with your internal workflows.
The company claims that early results show that RosettaStoneLLM can perform tasks akin to entity mapping and code generation as much as 30 percent higher than general models of GPT-4.
An organization spokesperson told VentureBeat: “In the company world, particularly in industries like insurance and finance, structured data is critical to operations.” Imagine an insurance company making a medical health insurance quote for a customer. The customer submits their HR database information, which is commonly in various formats and structures, with unique naming conventions for columns, rows and abbreviations. Traditionally, converting this data right into a format compatible with the insurance company's internal analytics systems is time-consuming. However, RosettaStone can accomplish this transformation in only just a few seconds, a task that may otherwise take man-hours.”
Hercules AI, based in Campbell, California, has raised $12.1 million in enterprise capital and, while it didn't disclose client numbers, said it was supported by Fortune 1000 firms and 30 percent of the highest U.S. law firms is getting used
