The CEL Framework - Using AI for Personal and Business Use
By: Cameron Duncan (originally posted as a LinkedIn Article)
After a years' worth of leading workshops, training clients, and building AI-powered tools, I found myself on a mission: to find a practical, straightforward framework to communicate the best practices of working with large language models (LLMs). LLMs - with ChatGPT being the most well-known example - are advanced AI systems designed to understand and generate human-like text based on the data they’ve been trained on.
I spent a lot of time searching, reading, and testing different approaches.
But over and over, I ran into the same issues - most frameworks were either so complex they felt overwhelming, so generic they weren’t actionable, or only really useful for personal projects. But my work is focused on how businesses can leverage this technology, and nothing fit perfectly.
So, I decided to create my own. Over the past six months, I’ve been refining this framework through real-world experience - workshops, client engagements, and plenty of hands-on experimentation. Now, I’m excited to share it more broadly.
CEL stands for Capabilities, Enablers, and Limiters. My goal with this framework is to make working with LLMs both effective and accessible—something you can actually use, whether you’re an individual contributor or leading a team.
But Cameron, why another framework?
Yes, there are plenty of AI frameworks out there, but most are either too technical for most people to use or too vague to be helpful. What’s missing is a practical approach that balances depth with simplicity—something that helps you get real value from LLMs, whether you’re using them for business or personal projects.
That’s where CEL comes in. It’s easy to remember, but comprehensive enough to guide you through the real challenges and opportunities of working with LLMs.
First, let’s look at what LLMs are actually good at. I think of them as versatile digital assistants—capable of handling a wide range of tasks. Here are the six main capabilities:
Summarize: LLMs can take large amounts of information and distill it into clear, concise summaries. Whether you’re dealing with lengthy reports or a flood of meeting notes, they can help you focus on what matters most.
Translate: Beyond translating between languages, LLMs can adjust tone, complexity, and style. Need a technical explanation made simple, or a casual note made more formal? LLMs can do that.
Analyze & Classify: LLMs can interpret data, spot patterns, and sort information into categories—even when things aren’t black and white. From sentiment analysis to document tagging, they bring a new level of nuance to data analysis.
Retrieve: If you have a question, LLMs can often provide a relevant answer—quickly and accurately—based on their training data.