A New Era of Data Reconciliation Using LLMs | SmartStream

AI has the power to revolutionise reconciliation.

SmartStream‘s platform AIR has been designed to do just that, streamlining financial data processing with advanced automation and intelligent matching. 

CTO Andreas Burner fills us in on new features such as enhanced cash and Nostro reconciliations, and looks at how large language models can be integrated into a tech stack to enable real-time, conversational data analysis.

Watch the video to find out:

• How AIR’s Affinity component boosts manual reconciliation by up to 30% for large institutions.
• The power of scalability in software.
• More info on an exciting practical use case for conversational AI. 

AI is increasingly being used across financial services to streamline data reconciliation processes and improve efficiency, accuracy, and adaptability. The role of AI in reconciliation is often cited as a developing trend in the industry.

Advancing Reconciliation with SmartStream AIR

In this conversation, SmartStream CTO Andreas Burner tells us that AIR, now in its ninth iteration, has undergone significant improvements, particularly in user onboarding and workflow optimization. The platform’s user interface (UI) and user experience (UX) have also been refined to ensure a more seamless experience. 

The standout feature of AIR 9 it would seem is the introduction of domain models. Given SmartStream’s expertise in cash, securities, reconciliations, and confirmations, these domain models enhance AIR’s ability to intelligently reconcile cash and Nostro accounts. This added intelligence enables financial institutions to manage reconciliations more effectively, reducing manual effort and increasing accuracy, one of the major goals in the industry today.

Harnessing AI for Better Insights

AIR has a number of other AI-driven components, each designed to enhance data processing and reconciliation. One key feature Burner points out is AI-assisted onboarding, which automatically recognizes and processes various data formats—structured, semi-structured, and unstructured—helping institutions quickly integrate new data sources.

Another crucial AI capability is pattern recognition and data matching. Traditionally, these processes are labor-intensive and prone to human error. AIR’s AI tools analyze historical data to detect patterns and suggest matches, significantly reducing the time required for reconciliation while improving accuracy.

Large institutions processing vast amounts of financial transactions typically have teams dedicated to reconciliation. AIR Affinity, a feature within the platform, leverages automated rule-based matching alongside AI-assisted manual matching. By learning from historical data, it can suggest more accurate matches, improving manual reconciliation rates by up to 30%. This not only reduces operational costs but also enhances efficiency, making it a game-changer for high-volume financial operations.

Scalability and AI-Driven Exception Management

In the conversation we also find out about the scalable and adaptable nature of AIR’s cloud native architecture, ensuring it can cater to large and small companies. This elasticity is great for optimising costs. 

Looking ahead, Burner lets us know how SmartStream is exploring the integration of large language models (LLMs) into AIR’s functionality. We hear about a prototype AI agent that allows users to interact with their exception management data through natural language queries—similar to how ChatGPT operates. Instead of generating static reports, users can simply ask questions like, “What is my current risk exposure?” or “Generate a pie chart of outstanding reconciliations.” The AI agent then retrieves and visualizes the data in real time.

This is a big innovation and could change the game for how financial institutions access and analyze reconciliation data. They’re refining this technology in the coming months and years, so there’s no question of their commitment to AI driven automation. 

Catch our other interviews with SmartStream and more financial experts on our website.