Build custom financial forecasting apps with ChatGPT Work

Summary

For finance teams, closing the books is only the beginning, and the real work is giving leadership and the board a clear story in forecasting what comes next. By asking ChatGPT Work to turn the latest close into a site, users can include scenarios to test together. Sites provides the capability to create shareable, interactive experiences used to communicate complex ideas throughout the organization. The prompt requests building an interactive H2 FY26 finance review and forecast site using approved forecast and close results, comparing last quarter actuals and the latest outlook against the official plan, keeping proposed finance adjustments separate, enabling monthly and account level scenario testing with a chart and sliders to make adjustment, period switching, driver analysis, drill down from executive summary to close detail, and including an executive ready presentation view. ChatGPT Work loads Google Drive and NetSuite Finance integrations, reads files, and confirms the canonical blossom systems financial review source. The workflow requires raw reconciliation and a consistent ten percent presentation scale in the reader facing site so all totals and readiness logs remain at source precision while the dashboard displays executive scale values. The official close and outlook reconcile clearly, showing adjusted cloud revenue at thirty nine point five zero million dollars, and support at twelve point five two million dollars, while the supported FY26 opening is twelve point nine eight million dollars, and the residual risk is nine hundred twenty thousand dollars. The recommendation remains conditional because release gates are still outstanding, and the dashboard makes these gates and any ordered scenario impossible to confuse with posted actuals. The working model is built around the approved bridge and account level close evidence. The social preview requirement establishes that the published link has an executive ready finance card matching the finished review rather than a generic placeholder. The review features executive decision, period specific scenario cases, account contribution comparison, driver analysis, full close drill down, recovery actions, release gates, and a three slide presentation mode. Once published privately, the site shows an executive summary with a conditional go status, highlighting that the approved recovery restores ninety three point two percent of the contribution gap, and the release outlays remain after the remaining control and commercial thresholds. The official plan is thirty point nine zero million dollars, last quarter actual is seventeen point four six million dollars, latest approved outlook is twenty nine point nine eight million dollars, and residual risk is nine hundred twenty thousand dollars. The scenario lab allows adjusting assumptions and seeing how they change forecast trajectories, including contribution trajectory and quarter over quarter change. Drivers and close details show what is driving forecasts and who owns each follow up item. The board presentation includes four proofs before final go, covering post duplicate reversal with Joe Casson, secure nimbus evidence with Patrick Dennis, publish forecast refresh with Ho Joon, and issue final go with Jesse. Users can also run other scenarios, such as the ten ten scenario where partner contribution increases by ten percent and expenses decrease by ten percent, allowing teams to see how the strategy changes the forecast and whether it returns to plan.