Stop wasting hours on manual data cleanup. Learn how to use Copilot and agentic workflows to turn complex spreadsheet tasks into simple prompts.

We’ve moved into what people are calling the 'Agentic Era.' Instead of wrestling with complex VBA code or fragile scripts, you can just talk to your data in plain English.
Learn to create effective automations using excel and copilot ai


This framework is a method used to guide AI, like Microsoft Copilot, to produce accurate and high-quality results rather than generic responses. To use it, you provide a persona or role (Context), a specific action or business logic (Task), and a defined output style or structure (Format). For example, instead of saying "summarize this," you would tell the AI to act as a "Senior Data Analyst," create a specific new column based on a logic rule, and format the output as a conditional formula.
Formatting your data as an official Table using the Ctrl+T shortcut is essential because it makes the data "AI-ready." Without these defined boundaries, the AI is essentially flying blind and may struggle to identify where headers and data ranges begin or end. Once the data is in a Table, the AI can recognize column headers as specific variables, which allows for much more precise manipulation and calculation.
While formulas are useful for answering a single question or calculating a specific value, scripts are designed to automate entire processes. A formula is like a snapshot, whereas a script is like a movie that can perform a sequence of actions, such as hiding columns, generating pivot tables, and saving files as PDFs. In the modern era of Excel, users can use natural language to ask Copilot to generate Office Scripts, allowing them to build tools that perform repetitive tasks forever rather than just calculating a result once.
To ensure accuracy, you should use a "Review and Refine" step by asking the AI to explain the logic behind the formula or script it generated. If the explanation does not match your business requirements, the logic is likely incorrect even if the formula returns a value. Additionally, you can build a "Safety Net" by including a validation step in your prompt, such as asking the AI to provide a summary count of how many rows were modified to verify the output against your expectations.
"Few-Shot" prompting is a technique where you provide the AI with a specific example or pattern to follow within your prompt. By showing the AI one "Before" and "After" example—such as how a phone number should be formatted—you eliminate guesswork and help the AI understand the exact transformation you want. This is often the fastest way to get precise formatting or data cleanup results without needing to write complex technical instructions.
From Columbia University alumni built in San Francisco
"Instead of endless scrolling, I just hit play on BeFreed. It saves me so much time."
"I never knew where to start with nonfiction—BeFreed’s book lists turned into podcasts gave me a clear path."
"Perfect balance between learning and entertainment. Finished ‘Thinking, Fast and Slow’ on my commute this week."
"Crazy how much I learned while walking the dog. BeFreed = small habits → big gains."
"Reading used to feel like a chore. Now it’s just part of my lifestyle."
"Feels effortless compared to reading. I’ve finished 6 books this month already."
"BeFreed turned my guilty doomscrolling into something that feels productive and inspiring."
"BeFreed turned my commute into learning time. 20-min podcasts are perfect for finishing books I never had time for."
"BeFreed replaced my podcast queue. Imagine Spotify for books — that’s it. 🙌"
"It is great for me to learn something from the book without reading it."
"The themed book list podcasts help me connect ideas across authors—like a guided audio journey."
"Makes me feel smarter every time before going to work"
From Columbia University alumni built in San Francisco
