The New York-based startup Cognaize has raised $18 million to develop its unstructured data processing platform for use in financial artificial intelligence (AI) applications. The company takes a hybrid approach, utilizing both deep learning technology that has been trained on financial models and human input to hone the final product. Cognaize will use the funds to increase staff size, launch new research and development efforts, and create additional products. Argonautic Ventures served as the round’s lead investor, joined by Metaplanet and other backers.
Although AI has had far-reaching effects in the banking industry, the majority of current efforts are focused on structured data. Cognaize’s platform is designed to take advantage of the tidal wave of unstructured data that is waiting to be mined for useful insights. The company operates under the assumption that despite the vast amounts of data available to the financial sector today, only a small fraction of that data (the structured part) is actually used to improve understanding of services, market conditions, and customer preferences. Cognaize uses deep learning that has been trained on financial models and a large corpus of documents to remedy this problem.
Cognaize’s infrastructure is based on deep learning tools that have been trained exclusively on financial models. The 1.3 million documents accessed by the platform could contain numerous “cells” of information that would require a trained eye to “read.” Loan applications, SEC filings, environmental, social, and governance (ESG) documents, presentations, trustee reports, and others fall under this category.
The company has developed a platform that draws from deep learning trained on financial models and a wide range of documents (over 1.3 million in total) that may contain numerous “cells” of information that require a more trained eye to “read.” (Not only do the documents include SEC filings, ESG documents, presentations, trustee reports, and more, but they also cover loan applications.)
Cognaize acknowledges that AI has the potential to advance in ways never before imagined, but it will never be able to fully replace humans. Therefore, the platform also includes “humans in the loop” to help fine-tune the work. Financial analysts and other human workers use the platform to correct the readings and draw conclusions and decisions based on the results.
Cognaize’s founder, CTO, and Chief Product Officer is Vahe Andonians. A prior fintech he founded, which focused on analytics and risk management for credit investments, was acquired by Moody’s. His philosophy there and with Cognaize is based on the idea that AI can complement human beings by performing tasks that humans cannot.
Two of the three largest credit rating agencies, as well as major insurance providers and banks, are among the many large companies that have signed on as Cognaize clients. Customers can process massive amounts of unstructured financial data and extract insights with remarkable precision and speed using the company’s platform, leading to better decision-making, risk assessment, and the discovery of patterns and trends that were previously obscured by complexity and human error.
The success of AI-focused startups like Cognaize exemplifies a key trend in the industry. An interesting trend is the rise of very strong players specializing in specific fields and use cases, even as many companies make big swings at general knowledge graphs in AI and create truly “large” large language models. It’s possible that these powers are still constructing “large” LLMs, but they’re putting more emphasis on quality than breadth.
It’s possible that investors are betting on specialists because they believe they will always be able to communicate with their customers on a more personal level in the language they use. Better results and training on the company’s unique needs are the strongest arguments for more tailored approaches, but they may also be less expensive to run because their LLMs have fewer parameters and thus require less processing power.
Viken Douzdjian, managing Partner at Argonautic Ventures, said in a statement, “We are thrilled to partner with Cognaize as they apply the transformative power of AI and large language models (LLMs) to finance.” Although AI has caused disruption in many sectors, there are countless applications for finance-specific generative AI due to the vast quantities of unstructured financial data. Our faith in Al, Vahe, and the rest of the Cognaize team to shape the way AI is used in finance is solid.
In summary, Cognaize is an exciting startup to keep an eye on because they are one of the first to provide consistent, quantifiable benefits from artificial intelligence in the banking sector. For Rauno Miljand, managing partner at Metaplanet, investing in Al, Vahe, and the rest of the Cognaize team was a no-brainer. “They have already harnessed the power of AI, as evidenced by the enviable expansion of Cognaize’s business, the global leaders in finance that they have secured as customers, and their unmatched technology roadmap. They are rapidly redefining the limits of what modern AI can do for the financial sector by leveraging the power of their own data to slash expenses dramatically and gain a competitive edge.
First reported on TechCrunch
Frequently Asked Questions
Q. What is Cognaize, and what is its recent funding announcement about?
Cognaize is a New York-based startup focused on developing an unstructured data processing platform for financial artificial intelligence (AI) applications. The company recently raised $18 million in funding to further develop its platform, which combines deep learning technology with human input to extract useful insights from the vast amounts of unstructured data in the financial sector.
Q. What sets Cognaize’s platform apart from other AI efforts in the banking industry?
Unlike many AI initiatives that primarily focus on structured data, Cognaize’s platform is designed to leverage the enormous potential of unstructured data in the financial sector. It operates under the belief that only a small fraction of available data is currently used to improve services, market understanding, and customer preferences.
Q. How does Cognaize’s platform process unstructured financial data?
Cognaize’s platform is built on deep learning tools trained on financial models. It analyzes over 1.3 million documents that may contain valuable information, such as loan applications, SEC filings, ESG documents, presentations, and trustee reports. These “cells” of information require a trained eye to interpret accurately.
Q. How does Cognaize incorporate human input into its platform?
Recognizing that AI cannot entirely replace humans, Cognaize includes a “humans in the loop” approach. Financial analysts and other human workers utilize the platform to fine-tune the readings, draw conclusions, and make decisions based on the results.
Q. Who is Vahe Andonians, and what is his vision for AI in finance?
Vahe Andonians is the founder, CTO, and Chief Product Officer of Cognaize. He believes that AI can complement human capabilities by performing tasks beyond human abilities. His prior fintech venture, which focused on analytics and risk management for credit investments, was acquired by Moody’s.
Q. Which industries and companies are utilizing Cognaize’s platform?
Cognaize’s clients include two of the three largest credit rating agencies, major insurance providers, and banks. These companies use the platform to process vast amounts of unstructured financial data, leading to better decision-making, risk assessment, and discovery of hidden patterns and trends.
Q. Why are investors interested in AI-focused startups like Cognaize?
Investors recognize the potential of specialists like Cognaize, who focus on specific fields and use cases in AI. Tailored approaches offer better results and training for a company’s unique needs and may be less expensive to run with fewer parameters in their large language models.
Q. What do the lead investors, Argonautic Ventures and Metaplanet, have to say about Cognaize?
Argonautic Ventures believes in the transformative power of AI and large language models (LLMs) in finance and has faith in Cognaize’s team to shape AI’s future in the financial sector. Metaplanet, another investor, praises Cognaize for harnessing the power of AI and its unmatched technology roadmap, leading to significant cost reductions and a competitive edge in the financial industry.
Q. Why is Cognaize considered an exciting startup in the banking sector?
Cognaize stands out for providing consistent, quantifiable benefits from artificial intelligence in the banking sector. Its platform’s ability to unlock insights from unstructured data sets it apart in the industry, making it an innovative company to watch closely.
Q. What does the future hold for Cognaize and its platform?
Cognaize’s future roadmap involves further development and expansion of its technology, enabling even more applications and benefits for the financial sector. The company’s AI-driven approach is poised to reshape how data is utilized in finance and bring valuable insights to businesses and decision-makers in the industry.
Featured Image Credit: Unsplash