--- title: How do I flag unusual utility bills across my buildings, adjusted for weather, with AI? | AEC Agents url: https://aecagents.si/use-cases/utility-bills-weather-normalized-flags description: An owner with several buildings gets a stack of electric, gas and water bills every month in different formats, and the real problems hide in them: an estimated read that will catch up later, a slow leak that adds a few hundred dollars a month, a boiler or air handler running at night. Comparing this month with last month mostly flags the weather. Have Claude or ChatGPT read the bills into one table, and have Copilot in Excel or Gemini in Google Sheets compare each bill with the same building's history after adjusting gas for heating degree days and electricity for cooling degree days, and flag what's left. What's actually wrong is for your facilities team to find on site. last_modified: 2026-11-11 source: AEC Agents (free, open workflows; not engineering advice) --- [AI assists](https://aecagents.si/use-cases/utility-bills-weather-normalized-flags#honesty-labels)Workflow guide · not independently tested # How do I flag unusual utility bills across my buildings, adjusted for weather, with AI? An owner with several buildings gets a stack of electric, gas and water bills every month in different formats, and the real problems hide in them: an estimated read that will catch up later, a slow leak that adds a few hundred dollars a month, a boiler or air handler running at night. Comparing this month with last month mostly flags the weather. Have Claude or ChatGPT read the bills into one table, and have Copilot in Excel or Gemini in Google Sheets compare each bill with the same building's history after adjusting gas for heating degree days and electricity for cooling degree days, and flag what's left. What's actually wrong is for your facilities team to find on site. [Owner](https://aecagents.si/roles/owner) · Owner & development · Claude · ChatGPT · Microsoft Excel · Microsoft Copilot · Google Sheets Content updated 11 November 2026 · Tool links listed below; vendor features and pricing change, so confirm with the vendor ## Short answer Use Claude or ChatGPT, then Copilot in Excel or Gemini in Google Sheets: Claude or ChatGPT reads every utility bill into one table with read type and service dates; Copilot or Gemini fits each building's history against degree days and flags bills that stay unusual after the weather is accounted for. An owner with several buildings gets a stack of electric, gas and water bills every month in different formats, and the real problems hide in them: an estimated read that will catch up later, a slow leak that adds a few hundred dollars a month, a boiler or air handler running at night. Comparing this month with last month mostly flags the weather. Have Claude or ChatGPT read the bills into one table, and have Copilot in Excel or Gemini in Google Sheets compare each bill with the same building's history after adjusting gas for heating degree days and electricity for cooling degree days, and flag what's left. What's actually wrong is for your facilities team to find on site. A person still checks: every flag on site with the facilities team before a bill is disputed or a repair is ordered. ## AI tie-in **Tool:** Claude or ChatGPT, then Copilot in Excel or Gemini in Google Sheets Claude or ChatGPT reads every utility bill into one table with read type and service dates; Copilot or Gemini fits each building's history against degree days and flags bills that stay unusual after the weather is accounted for. ### Prompt 1 ``` Read these utility bills [upload]. Build a table with one row per bill: building, utility, meter, service start and end, usage, unit, amount, demand if shown and actual or estimated read. Copy figures exactly and write NOT SHOWN instead of guessing. List bills whose service period overlaps or leaves a gap with the previous one. ``` ### Prompt 2 ``` Here are the headers and a few sample rows of my Bills and Weather tabs [paste]. Write Excel formulas (or a step-by-step Copilot plan) that fit gas against heating degree days and electricity against cooling degree days per building, compute expected use and the difference for each new bill, and flag differences over [percent], estimated reads and true-ups, water above prior years, demand spikes and period gaps. Use only my data and don't guess causes. ``` **Human check:** Every flag on site with the facilities team before a bill is disputed or a repair is ordered. ## The problem Flag unusual utility bills across your buildings, adjusted for weather, with AI. This is a common, time-consuming task for a owner. ## What this workflow does An owner with several buildings gets a stack of electric, gas and water bills every month in different formats, and the real problems hide in them: an estimated read that will catch up later, a slow leak that adds a few hundred dollars a month, a boiler or air handler running at night. Comparing this month with last month mostly flags the weather. Have Claude or ChatGPT read the bills into one table, and have Copilot in Excel or Gemini in Google Sheets compare each bill with the same building's history after adjusting gas for heating degree days and electricity for cooling degree days, and flag what's left. What's actually wrong is for your facilities team to find on site. **How it works:** AI drafts or analyzes. A person operates the tools, checks the work, and takes action. **Your checkpoint:** The flags only reflect the bills, weather data and thresholds you entered. A simple degree-day fit misses changes in occupancy, hours or equipment, so some flags will be explained by things the spreadsheet can't see, and some real problems won't stand out. Disputing a bill, repairs and any energy reporting stay with your facilities team and whoever signs off on them. ## Before you start - Access to the named tool; confirm license, version and cost with the vendor. - A small, non-confidential sample to try the workflow on first. - A qualified person to review the output before it is used. ### Tools and access - [Claude (Anthropic)](https://www.anthropic.com/claude) - [ChatGPT](https://chatgpt.com/) - [Microsoft Excel](https://www.microsoft.com/microsoft-365/excel) - [Microsoft Copilot](https://www.microsoft.com/microsoft-365/copilot) - [Google Sheets](https://workspace.google.com/products/sheets/) ## Follow the steps 1. Collect the last two or three years of utility bills or the utilities' data exports for each building, and the monthly heating and cooling degree days for the nearest weather station from a public weather source. 2. Have Claude or ChatGPT read every bill into a Bills tab: building, utility, account or meter, service start and end dates, usage, unit, amount, demand if shown, and whether the read was actual or estimated. It copies figures exactly and writes NOT SHOWN instead of guessing. 3. Have it convert units consistently (for example all gas in one unit) and list any bill with a service period that overlaps or leaves a gap with the previous one. 4. Put the degree days in a Weather tab, and have Copilot in Excel or Gemini in Google Sheets fit each building's gas use against heating degree days and electricity against cooling degree days from the history, then compute expected use for each new bill and the difference in units and dollars. 5. Flag bills more than a threshold you set from expected, estimated reads and the true-up bills after them, water use above the same month in earlier years, demand spikes, rate changes, and bills with gaps or overlaps. 6. Send the flags to the facilities team with the likely cause written as a question to check (meter, controls, leak, schedule change, billing error) and track what they found next to each row. ## Try a small example Replace the bracketed fields with your own project details. No tested sample run is published for this workflow yet. ### Sample input ``` Fill in the bracketed [fields] in the text below with your project data. ``` ### Starter prompt ``` Here are the headers and a few sample rows of my Bills tab [paste: building, utility, meter, service start, service end, usage, unit, amount, demand, read type] and my Weather tab [paste: month, heating degree days, cooling degree days]. Write Excel formulas (or a step-by-step Copilot plan) that fit each building's monthly gas use against heating degree days and electricity against cooling degree days from the history, compute expected use and the difference in units and dollars for each new bill, flag differences over [percent], estimated reads and the next bill after them, water above the same month in prior years, demand spikes, rate changes and service period gaps or overlaps, and summarize by building. Use only my data and don't guess the cause. ``` ### Expected output ``` Not published yet. Treat any output as a draft and check it against the source documents. ``` ## Check the result - Every value, citation or quantity traces back to a source document. - Missing information is marked as missing, not filled in. ## Limits and recovery - The flags only reflect the bills, weather data and thresholds you entered. A simple degree-day fit misses changes in occupancy, hours or equipment, so some flags will be explained by things the spreadsheet can't see, and some real problems won't stand out. Disputing a bill, repairs and any energy reporting stay with your facilities team and whoever signs off on them. - Check your organization’s data policy before sharing project files with third-party tools. **Stop or switch when:** The source cannot be verified, permissions are missing, or the result contains unsupported facts. ## Frequently asked questions How does AI help with this? Claude or ChatGPT, then Copilot in Excel or Gemini in Google Sheets: Claude or ChatGPT reads every utility bill into one table with read type and service dates; Copilot or Gemini fits each building's history against degree days and flags bills that stay unusual after the weather is accounted for. A person still checks: Every flag on site with the facilities team before a bill is disputed or a repair is ordered. How do I how do I flag unusual utility bills across my buildings, adjusted for weather, with AI?? Collect the last two or three years of utility bills or the utilities' data exports for each building, and the monthly heating and cooling degree days for the nearest weather station from a public weather source. Have Claude or ChatGPT read every bill into a Bills tab: building, utility, account or meter, service start and end dates, usage, unit, amount, demand if shown, and whether the read was actual or estimated. It copies figures exactly and writes NOT SHOWN instead of guessing. Have it convert units consistently (for example all gas in one unit) and list any bill with a service period that overlaps or leaves a gap with the previous one. The full step list is above. Does this use an AI agent or regular software? It is labeled AI assists: AI drafts or analyzes. A person operates the tools, checks the work, and takes action. What tools do I need? Claude, ChatGPT, Microsoft Excel, Microsoft Copilot, Google Sheets. Confirm license, version and cost with the vendor, and try it on a small non-confidential sample first. Has this workflow been tested? Workflow guide · not independently tested. Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written. ## Where to go next - [How do I compare site captures to the plan with AI progress tracking?](https://aecagents.si/use-cases/progress-tracking) - [How do I capture existing conditions and as-builts with 3D scanning?](https://aecagents.si/use-cases/as-built-capture) - [How do I review quality and progress remotely with AI-analyzed site captures?](https://aecagents.si/use-cases/remote-quality-control) - [All Owner workflows](https://aecagents.si/roles/owner) - [All Schedules, submittals and project controls workflows](https://aecagents.si/tasks/project-controls) ## Evidence and maintenance Written from vendor documentation and practitioner use. Labels apply to the steps described here, not every feature of a product. Links were checked when the page was written. - [Claude (Anthropic) (vendor site)](https://www.anthropic.com/claude) - [ChatGPT (vendor site)](https://chatgpt.com/) - [Microsoft Excel (vendor site)](https://www.microsoft.com/microsoft-365/excel) - [Microsoft Copilot (vendor site)](https://www.microsoft.com/microsoft-365/copilot) - [Google Sheets (vendor site)](https://workspace.google.com/products/sheets/) ## Put it to work Start with a small sample and keep the original record open while you verify the result. ## Know what does the work. Agent executes : An AI agent calls connected tools to carry out bounded actions. You set permissions and review the result. AI assists : AI drafts or analyzes. A person operates the tools, checks the work, and takes action. Conventional automation : Rules or specialist software perform the task. No AI agent is required.