Extracting valuable insights from unstructured text is a critical application in the finance industry. However, this task often goes beyond simple data extraction and necessitates advanced reasoning capabilities. A prime example is determining the maturity date in credit agreements, which usually involves deciphering a complex directive like “The Maturity Date shall fall on the last Business Day preceding the third anniversary of the Effective Date.” This level of sophisticated reasoning poses challenges for Large Language Models (LLMs). It requires the incorporation of external knowledge, suc ..read more

Michael Hughes, the chief business officer for Duality Technologies, could barely conceal his excitement when the White House Office of Science and Technology issued a directive calling for the adoption of privacy-enhancing technology recently. “I was just blown away,” Hughes said. “That was big.” Up to this point, the market at large has been relatively slow to adopt privacy-preserving technologies, which allow sensitive data to be used for analytics or AI without sacrificing privacy. Hughes expects that to change soon in light of several developments, including a potential new US law dubbed ..read more

Snowflake today took the wraps off Arctic, a new large language model (LLM) that is available under an Apache 2.0 license. The company says Arctic’s unique mixture-of-experts (MoE) architecture, combined with its relatively small size and openness, will enable companies to use it to build and train their own chatbots, co-pilots, and other GenAI apps. Instead of building a generalist LLM that’s sprawling in size and takes enormous resources to train and run, Snowflake decided to use an MoE approach to build an LLM that’s smaller than massive LLMs but can offer a similar level of language unders ..read more

The process of annotating and labeling data is critical for supervised learning tasks, such as training a large language model (LLM) and other types of machine learning models. However, the need for human cognition and input is a limiting factor on the amount of data that can be prepared. As a result, there is considerable demand for software that can help streamline the data labeling and annotation workflow, as well as for third parties that can do the labeling work on an outsourcing contract. Everest Group recently ranked the top providers in these booming spaces. Everest Group, a Dallas, Te ..read more

EnterpriseDB next month is expected to formally launch a new lakehouse that puts Postgres at the center of analytics workflows, with an eye toward future AI workflows. Currently codenamed Project Beacon, EDB’s new data lakehouse stack will utilize object storage, an open table format, and query accelerators to enable customers to query data through their standard Postgres interface, but in a highly scalable and performant manner. The popularity of Postgres has skyrocketed in recent years as organizations have widely adopted the open source database for new applications, especially those runnin ..read more

Altair Engineering will soon fortify its end-to-end data science offering with the acquisition of Cambridge Semantics and its semantic graph database. Terms of the deal, which was announced yesterday, were not disclosed. Cambridge Semantics was founded in 2007 by a group of researchers from IBM’s Advanced Technology Internet Group with the goal of using emerging semantic Web technologies to make data more searchable and easier to query. From its headquarters in Boston, Massachusetts, the company developed a graph database dubbed Anzo that was based on emerging Resource Description Framework (R ..read more

Steve Jobs once compared the electric motor to the democratization of new technologies. When the electric motor was first invented, it could only be manufactured on a large-scale in a factory. They had great applications, but required significant resources to maintain, from elaborate belts and pulleys to a team of skilled mechanics. It was elusive to create a true lasting impact. Somewhere along the way, these large electric motors were downsized to fractional horsepower electric motors and proliferated to the point where we now easily have over 50 of them in the average household. We’re at a ..read more

Large language models (LLMs) hallucinate. There’s just no way around that, thanks to how they’re designed and fundamental limitations on data compression as expressed in the Shannon Information Theorem, says Vectara CEO Amr Awadallah. But there are ways around the LLM hallucination problem, including one by Vectara that uses retrieval-augmented generation (RAG), among other methods. MIT professor Claude Shannon, who died in 2001, is known as the Father of Information Theory thanks to his extensive contributions to the fields of mathematics, computer science, electrical engineering, and cryptog ..read more

Informatica, one of the preeminent names in ETL, is considering a deal to be acquired by enterprise cloud software provider Salesforce, according to a report by the Wall Street Journal last week. If completed, the deal would net Salesforce a large catalog of well-respected data management tools and a customer list full of blue-chip names. According to the WSJ’s April 12 story, Salesforce is in “advanced talks” to acquire Informatica, which would be among its largest acquisitions to date. While a price wasn’t specifically mentioned, the WSJ says the price its anonymous source disclosed is below ..read more

The Collibra community descended upon Orlando, Florida, this week for the company’s annual Data Citizens conference. In addition to educational sessions and keynotes on the importance of data intelligence, attendees got their first glimpse of Collibra AI Governance, the company’s latest product, which is aimed at helping customers get more control over the AI models they put into production. Companies around the world are rushing to develop AI solutions to improve their operations in a variety of ways. Generative AI, in particular, has caught the attention of executives and board members who p ..read more