Guozhen AIGlobal AI field notes and model intelligence

Realtime AI News

Northern Trust and Microsoft deploy an AI agent that cuts onboarding prep from weeks to days

Northern Trust has expanded its Microsoft collaboration with an AI agent that prepares clients' historical investment and transaction data for onboarding on its Front Office Solutions platform, cutting data preparation time by 83%, from about two and a half weeks to three days. The agent runs on Azure and Microsoft Foundry, integrates with Excel and SharePoint, and reports a 96.5% first-pass F1 quality score, with analysts reviewing its output.

Published

Northern Trust has expanded its collaboration with Microsoft, starting with an AI agent that prepares clients' historical data for onboarding. According to The Paypers, the agent was developed by Northern Trust and Microsoft Frontier Company's Forward Deployed Engineering teams.

It is being used for clients of Northern Trust Asset Servicing's Front Office Solutions platform, which provides performance and analytics for institutional investors such as foundations, endowments, pension schemes and large family offices.

The work itself is unglamorous but heavy: converting years of a client's investment and transaction data into the standardized format used in Northern Trust's reporting environment. Onboarding can require consolidating large volumes of data from multiple sources to reconstruct performance histories, particularly for alternative assets. In a Northern Trust survey, 57% of asset owners cited consolidating data from disparate internal and external sources as their biggest challenge.

The results are concrete. Northern Trust says the agent cuts data preparation time by 83%, from two and a half weeks to three days, and delivers a 96.5% first-pass F1 quality score, a measure that combines precision and recall.

Humans stay in the loop. Analysts review the agent's outputs, evaluate its reasoning, resolve exceptions and feed improvements back into it; during implementation the agent is used to reconcile errors and validate a client's starting data before the client goes live.

The agent runs on Microsoft Azure and Microsoft Foundry and integrates with Excel and SharePoint, while an Azure-based intelligence layer preserves approved onboarding knowledge so the process improves over time.

On governance, Microsoft tuned the agent's quality using only anonymised data, so Northern Trust client data remained under the firm's control. Pete Cherecwich, COO of Northern Trust, said the agent combines the firm's asset servicing expertise with Microsoft's AI capabilities to get clients up and running more quickly while maintaining the oversight and controls clients expect.

Bill Borden, Corporate Vice President, Worldwide Financial Services at Microsoft, said the collaboration shows how AI can amplify human expertise to speed up service delivery with governance and human oversight built in, and that the approach could apply to complex workflows across regulated industries.

Northern Trust, founded in 1889, reported USD 19.9 trillion in assets under custody or administration and USD 2.0 trillion in assets under management at the end of September 2026. For banks, back-office onboarding work like this is where agents produce measurable returns first, and it is becoming a common entry point for AI moving from pilots into production.

Why it matters

The case matters because it puts agent ROI into auditable numbers: weeks to days, plus a 96.5% first-pass quality score. If peers can reproduce those figures, procurement of agents inside regulated bank back offices shifts from experimental budgets to efficiency programs.

MicrosoftAI AgentFinance
Back to realtime news

Nearby Updates

All

11/13, 01:00

Phota Labs launches an iOS app that re-shoots everyday photos with AI

Phota Labs has released an iOS app that uses AI to turn everyday snapshots into professional-looking photographs while keeping people's real appearance intact. The a16z-backed startup, founded by two former Adobe computational photography engineers, pitches it as a photographer that works before you ask.

11/13, 01:04

Trust Stamp adds liveness checks so humans approve AI-agent transactions

Trust Stamp has launched Knowing Consent, a biometric approval layer that verifies a bank customer in person before a transaction initiated by an AI agent is authorized. The system separates an agent's access to an account from the account holder's approval of one specific payment.

11/13, 01:04

Bielik 11B v3.1 ships as a 56-language Polish open model

Poland's Bielik 11B v3.1 is out with support for 56 languages, widening the multilingual footprint of Europe's open-weight model family. The instruction-tuned release ships under the Apache 2.0 license with GGUF, AWQ and MLX quantized builds intended for local and private deployment.

11/13, 00:12

After Huang's resignation, Trump floats Tim Cook to run Nvidia

President Trump posted on Truth Social that only one man could replace Jensen Huang as Nvidia's chief executive: Apple's Tim Cook, writing that “A Legend replaces a Legend!” The suggestion follows Huang's resignation on Thursday and a Time op-ed by Yale's Jeffrey Sonnenfeld recommending Cook, though CEO succession at Nvidia is a board decision rather than a presidential one.