- The Growth Ledger
- Posts
- Smarter Tax Workflows: AI as Your Next Best Hire
Smarter Tax Workflows: AI as Your Next Best Hire
Success comes from treating AI like a highly capable team member who needs both guidance and oversight
Edna Garcia
February 27, 2025 • Estimated Reading Time: 5 minutes
Hey — It's Edna.
Welcome to this week's edition of The Growth Ledger, your resource for modernizing your tax practice.
"Think of AI like your most capable staff member—but one who needs clear instructions and always requires review."
When Orumé Hays, CPA speaks about AI in tax practice, people listen. And for good reason. Beyond her role as a statewide Director at NYSSCPA and chair of the Relations with the IRS Committee, she's built a reputation for turning complex tech concepts into practical solutions for tax professionals. Her latest insight cuts through the AI hype with a simple truth: success comes from treating AI like a highly capable team member who needs both guidance and oversight.
That shifts the entire conversation from "Will AI replace us?" to "How do we make it work for us?"
Today, we'll dive into:
How to actually use AI in your daily tax work
Real examples from leading practitioners
A practical framework for implementation
Clear steps for getting started
Specificity brings Success for Tax Pros
Successful tax firms aren't trying to revolutionize everything at once with AI. They're identifying specific tasks, creating detailed guidelines, maintaining human oversight, and expanding capabilities gradually. This measured approach is delivering measurable results: 30% faster compliance processes and 40% fewer errors. |  |
The Current Opportunity
While 62% of marketing professionals are leveraging AI, only 37% of finance and accounting professionals have adopted these tools. For tax practitioners, this represents a significant competitive advantage if implemented correctly.
Three Implementation Strategies That Work
1. Craft Specific Prompts When Using ChatGPT and Similar AI Tools
Tax professionals getting valuable results from generative AI platforms like ChatGPT, Claude, and Microsoft Copilot have discovered the power of detailed prompt engineering. The specificity in your prompt creates guardrails that produce more precise, actionable guidance from generative AI. Unlike built-in tax software features, these tools require detailed context to generate reliable tax-related content. While generative AI excels at producing insights, AI agents like Numiro ensure those insights are integrated into workflows, streamlining execution and minimizing manual effort.
2. Start With Concrete Tasks
Focus AI implementation on areas with immediate ROI:
Retirement account tax analysis: Calculating RMDs, identifying missed opportunities, and optimizing withdrawal strategies
AI-Driven Tax Prep Integration: Numiro's OCR scans 1099s, 401(k) statements, and inherited IRA documents, auto-populating relevant fields in Drake, Ultratax and other tried and true tax software —eliminating manual entry errors in RMD calculations.
Client proposal drafting: Creating customized fee estimates based on complexity factors from prior years
Automated Sales Tax Compliance: Kintsugi monitors tax liabilities in real time, registers businesses in required states with a single click, and automates sales tax remittance—reducing errors and ensuring full compliance without the hassle.
Initial tax research: Identifying relevant code sections and recent precedents for specific client situations using Checkpoint Edge.
Financial pattern detection: Flagging unusual transaction patterns that might trigger correspondence audits
3. Implement Structured Review Processes
Treat AI like a junior associate with potential:
Document each AI recommendation with its supporting reasoning
Establish partner-level final review for all AI-assisted work
Track accuracy metrics to identify areas for improvement
Create client-facing language about how AI augments (not replaces) professional judgment. Look below for an example:
"Our AI tools act as tireless assistants, cross-checking 10,000+ data points across your filings. Every recommendation undergoes partner-level review to align with your unique financial goals. Think of it as having an entire research team verifying your return, guided by our CPAs' decades of experience." Wolters Kluwer CCH® ProSystem fx Engagement | Why this works so well:
|
Next Steps
Start by identifying one specific, bounded task where AI could support your existing workflow. Create detailed guidelines, implement oversight protocols, and measure results before expanding to additional areas.
A Reality Check
Here's what might surprise you: only 37% of all finance and accounting professionals are currently using AI, compared to 62% in marketing. That's not just a gap—it's an opportunity.
When we focus in on just tax practices, what’s clear is that those getting it right aren't trying to revolutionize everything at once. They're starting with specific tasks, maintaining clear oversight, and gradually expanding their AI capabilities.
As Hays emphasizes in her work with the NYSSCPA, success isn't about replacing judgment—it's about multiplying impact while maintaining control.
Join the Conversation
How are you thinking about AI in your practice? |
Until next week,
Edna Garcia
