Strategies and Solutions For Unlocking Value From Unstructured Data

How AI Transforms Unstructured Data into Strategic Assets for Financial Institutions: A-Team Webinar

In a world where 80-90% of enterprise data is unstructured, financial institutions face a dual challenge: harnessing its potential while navigating complexity, cost, and compliance. During A-Team Group’s recent webinar, Strategies and Solutions for Unlocking Value from Unstructured Data, industry leaders explored how AI is reshaping data strategies. 

The participants of the discussion  were:

• Vahe Andonians, Founder & CEO of Cognaize
• Victor Tewari, Senior Vice President - Wealth Management & Private Banking and Chief Data Officer, Citi
• Junaid Farooq, Senior Vice President, First Citizens Bank
• Nicole M. Allen, Director, Text Analytics, LSEG

Moderator: Mark McCord, Editor, Data Management Insight, A-Team Group

Discussion topics included tools required to extract data from unstructured sources, the value organisations can extract from unstructured data, integration challenges inherent in the data itself, technology-based solutions to successful management of unstructured data and others.

Andonians opened with a stark observation about the exponential growth of unstructured data:

“The amount of data today is humongous because we’ve all shifted from being mere consumers to producers - through social media, blogs, webinars, and more. As Eric Schmidt noted, humanity produced more data in the last two years than in all prior history - and most of it is unstructured.”

This shift has turned financial documents like loan agreements, KYC files, or syndicated loan notices into a tidal wave of untapped value. Yet, as Andonians stressed, the real challenge isn’t volume; it’s translating human-structured data into formats machines can analyze.

Furthermore, emphasized that the competitive edge in finance has fundamentally shifted: Twenty years ago, success came from exclusive data access. Today, it’s about who processes and applies data fastest. He echoed a modern revival of an old truth: “More data beats better models.” But this advantage only holds if organizations can filter out noise and extract actionable insights from the overwhelming volumes of unstructured content they handle daily.

Drawing on Google's concept of the "unreasonable effectiveness of data", Andonians illustrated how financial firms can move beyond one-size-fits-all solutions. By tapping into unstructured sources like emails, chat logs, or meeting transcripts, companies can deliver personalized, context-rich services - or “bespoke experiences.”

While generative AI introduces flexibility, it also brings unpredictability, called “stochastic process”:

“It hallucinates. That’s a feature, not a bug.”

To mitigate this, Cognaize uses hybrid validation systems: generative models extract data, while mathematical AI interpreters ensure outputs adhere to domain-specific rules (e.g., financial formulas). This marriage of creativity and control is key to scaling AI across risk-sensitive environments.

Watch the full recording of the webinar here. 

About Vahe Andonians

CEO & Founder, Cognaize

Vahe Andonians is a banker, data scientist, entrepreneur, and senior lecturer at the Frankfurt School of Finance & Management. Over the past 20 years, Vahe has been at the forefront of AI and deep learning, founding several AI start-ups, including an AI-based analytics company specializing in the fixed income market.