Northern Trust Deploys AI Agent to Accelerate Client Onboarding

Northern Trust and Microsoft announced on November 9, 2026 that a jointly developed client onboarding agent had reduced data preparation time by 83%, from about two and a half weeks to three days. The deployment begins an expanded AI collaboration and serves clients on Northern Trust Asset Servicing’s Front Office Solutions platform.

The system takes historical investment and transaction records spanning years and reformats them into Northern Trust’s standard data structure for reporting. Institutional clients such as pension schemes, endowments, foundations and large family offices use Front Office Solutions for investment performance and analytics. Reconstructing a new client’s performance history can require implementation teams to consolidate records from several sources, with alternative assets adding complexity.

Production Measurement and Quality Testing

Northern Trust measured the time saving in real client onboarding scenarios after deployment. The analysis covered data transformation, analyst review, exception handling and final validation, comparing the previous workflow’s baseline with the average completion time for agent-assisted production cases. The measurement therefore includes the surrounding analyst work rather than isolating the automated conversion step.

The companies also reported a 96.5% F1 first-pass quality score, using an evaluation metric that combines precision and recall. The test set comprised seven representative cases drawn from real historical client transaction workbooks that Northern Trust supplied in anonymized form.

Agent-generated values were compared with analyst-prepared reference outputs withheld during execution. The first-pass result was recorded before analyst feedback or corrections, so it measures the agent’s initial output rather than the quality after human review. That distinction separates the controlled quality test from the production timing, which includes analyst review and validation.

Azure Architecture and Human Control

Northern Trust developed the system alongside Microsoft Frontier Company Forward Deployed Engineering teams. Microsoft Foundry supplies the enterprise AI tooling, with Microsoft Azure providing the underlying platform. Integrations with Excel, SharePoint and other Microsoft 365 tools let analysts inspect, adjust and approve outputs. Azure also supports a governed intelligence layer that retains approved onboarding knowledge so the system can draw on institutional expertise over time.

Oversight continues through the workflow. Analysts can evaluate the agent’s reasoning, resolve exceptions and provide expert feedback. During implementation, before a client goes live, they can use the agent to identify and reconcile errors while validating the client’s starting data. This gives implementation teams a way to check the incoming history as well as the agent’s output.

Microsoft’s development process used what the companies call an “eyes-off” approach. Microsoft optimized F1 first-pass quality and other evaluation measurements using only anonymized data, while Northern Trust client data remained governed under the firm’s control. The arrangement was designed for financial-services privacy, security and data-stewardship requirements while allowing the joint engineering team to improve the system.

Northern Trust Chief Operating Officer Pete Cherecwich said the agent reduces the time needed to prepare and validate client data “while maintaining the oversight and controls our clients expect,” creating capacity for both sides to focus on higher-value work. Under the firm’s broader augmented-intelligence approach, AI extends human expertise while judgment and accountability remain with people.

The deployment ties a measurable reduction in preparation time to a specific division of responsibility: AI handles data conversion, analysts manage exceptions and validation, and Northern Trust retains control of client information. The operational question for broader use is whether the measured savings and first-pass quality remain consistent across varied client histories under those same controls.