Applied AI in finance: 4 tactics to automate advanced reporting
AI in finance series
Many finance professionals would agree they still dedicate too much time to creating reports. Three-quarters of finance teams spend 24 hours a month—three full workdays—just recreating financial reports.
For finance teams, much of AI's promise lies in its ability to handle the admin work—and let them get back to strategy. This second installment of our Applied AI in finance series shows how the latest tools can shrink reporting time by handling calculations, models, and narratives for you.
1. Calculate more advanced metrics
FP&A teams run models and projections to understand current performance and future trends. But compiling and manipulating this data takes time and often requires expert-level Excel skills. LLMs can make these insights accessible to a much larger group and give experts a boost by handling complex calculations that could take hours in spreadsheets.
Example: A retailer reviewing pricing wants to measure product elasticity. Nicolas Boucher, a veteran finance professional who currently runs AI Finance Club, demonstrated how you can upload a spreadsheet with item name, month, price, quantity sold, and total sales to ChatGPT, asking it to calculate elasticity. It used Python to run the equations and explained the formulas and results step by step, making them easy to verify.
2. Build ready-to-use financial models
GenAI doesn’t just calculate numbers—it can build dynamic, customizable financial models from scratch. You just describe what the model should show and the variables. Doing this in a spreadsheet would typically require a lot of trial and error (and perhaps a few hours watching tutorial videos).
Example: Continuing with the retailer, Boucher asked Anthropic’s Claude to model how store count affects profit, with adjustable variables for sales per store, margin, and fixed costs. Claude created three scenarios with different variables using React-based visual modeling. You can then download a version where parameters are adjustable and the graph updates accordingly that’s ready to share with others.
Want an Excel-ready version? Claude can walk you through the exact structure, references, and formulas—plus instructions for using Goal Seek, a what-if analysis tool in Excel, to find the optimal number of stores.
3. Draft executive-level narratives and commentary
Numbers are just the start. Finance leaders need to explain what’s driving budget vs. actual reports and month-over-month variance analysis in board and executive meetings. Writing these narratives from scratch takes time, especially for those who don't write often. GenAI can draft professional, tailored commentary that reflects their personal style to lighten the lift.
Example: Boucher gave ChatGPT a simplified P&L with actuals and budget, asking it to write board-ready commentary and noting that it should have specific figures to ensure those numbers are incorporated into the commentary. He illustrated how you can specify the desired tone (professional, actionable, direct) and format (bullet points, numbers-centric). ChatGPT's Canvas feature then makes it possible to adjust the output: change the reading level to length, give the work a final polish, and more. Canvas also lets you request specific changes like expanding a section or adding a source.
4. Turn these insights into decks
The final step is often turning data, models, and narratives into presentations. Getting slides to look just right usually takes longer than it should—but GenAI can help here, too.
Example: Once your narrative is ready, you can ask ChatGPT to create a PowerPoint file. Boucher showed how it generates a basic deck that can be styled with preloaded brand themes. You can also give design guidelines to generate something closer to the final product. Some refining is still necessary, but it gives you a strong start.
Looking for more ways to leverage AI?
Check back soon for more practical advice on how finance and accounting teams can drive value with AI.
For more tips on how your finance team can put AI into action, watch the full webinar with Boucher and join the 1 million+ people following him on LinkedIn.
