BPM needed AI that could scale across a growing firm without increasing risk or creating a patchwork of unmanaged tools. Over the past five years, the firm materially increased headcount and grew revenue by roughly 50%, which raised the bar on consistency, quality, and governance.

Instead of allowing ad hoc use of consumer AI tools, BPM chose to standardize on Microsoft 365 Copilot as its trusted, enterprise-grade platform because it:

• Embeds AI in the flow of work – directly in Word, Excel, Outlook, and Teams, so people don’t have to stop, switch tools, or relearn workflows.
• Aligns with existing security and governance – Copilot respects current roles and permissions, so people only see what they already have rights to access.
• Provides enterprise protections – data is not used for training, not retained, and is governed with identity and access controls that fit BPM’s Microsoft infrastructure.
• Supports firmwide licensing – everyone in the firm is licensed, giving colleagues a sanctioned AI experience instead of scattered, unmanaged tools.

Leaders also put an AI Acceptable Use Policy in place and required colleagues to read and acknowledge it. This combination of a single, secure platform plus clear guardrails allowed BPM to scale AI responsibly, without compromising on accuracy, trust, or compliance.

BPM uses Microsoft 365 Copilot and custom AI agents to help teams handle high-pressure, high-stakes work without lowering quality standards.

In busy season, Tax and Assurance teams face a compressed window where they must quickly rebuild client context—what happened last year, what changed, and what might affect filing positions. Copilot helps by:

• Summarizing client history and documents so teams can get oriented faster and reduce “startup friction.”
• Managing message overload by summarizing long email and Teams threads and surfacing what needs action.
• Refining communication – leaders report using Copilot “nine times out of ten” to fine-tune messaging, tone, and clarity before sending important emails.

For more complex needs, BPM uses AI agents that embed firm expertise directly into workflows:

• Tax planning agent – evaluates a tax return against BPM’s planning frameworks and surfaces potential opportunities earlier. This lets staff at earlier career stages take a first pass at planning and build analytical skills sooner.
• Quality checkpoint agent – applies clear criteria from subject matter experts to flag when work needs a second review from national tax or estate and trust specialists. This reduces guesswork and creates a documented record of due diligence.

Across the firm, leaders report that tasks that once took days or weeks can now be completed in a single day, while maintaining the same or higher quality. The goal is not to replace professional judgment, but to standardize the right questions, reduce friction, and free time for higher-value advisory work.

BPM treated Copilot as a change in how the firm works, not just another software rollout. The focus was on safe, responsible adoption at scale and on making AI genuinely useful in day-to-day work.

Key steps included:

• Firmwide licensing – BPM did not roll Copilot out in phases; they licensed everyone in the firm to avoid pockets of usage and shadow tools.
• Clear governance and policy – leaders created an AI Acceptable Use Policy, defined which AI tools were approved or restricted, and aligned Copilot with existing security, privacy, and governance requirements.
• Enterprise-grade security – with Microsoft, BPM ensured data is not used for training, not retained, and is protected by identity and access controls. Copilot respects existing roles and permissions, so users only see what they are already allowed to see.
• Role-based training and enablement – BPM partnered with Reply to deliver training tailored to specific roles and to show how Copilot fits into everyday workflows in Word, Excel, Outlook, and Teams.
• Ongoing governance program – IT treated Copilot as a firmwide governance initiative, putting guardrails, operating discipline, and adoption rhythms in place rather than a one-time deployment.
• Internal champions and practice leaders – with a relatively small IT team supporting about 1,300 people, BPM relied on practice leaders and internal champions to drive adoption and share practical use cases.

Leaders consistently reinforced that AI is there to remove friction, not replace judgment. For example, drafting policies and guidance that once took days to weeks now happens in a day, but the same review discipline and accountability for final work remain. This balance of enablement and control helped BPM reimagine how work gets done while maintaining trust, accuracy, and culture.