Kirsten Poon from Edmonton Shares 5 AI Lessons for Companies

Artificial intelligence is now part of how many companies operate. It shapes how work gets done and how decisions are made. Kirsten Poon from Edmonton has built AI systems for both commercial and industrial use. She has worked on projects that turn data into tools that solve specific problems. Through these projects, Kirsten Poon has learned what makes AI useful in real business settings. Here are five lessons she shares for companies that want to get real results from AI.

1. Start with a clear business problem

Some companies start AI projects without knowing exactly what they want to achieve. This often wastes time and resources. Kirsten Poon says you should always begin with a clear problem in mind. For example, a delivery company might want to reduce late shipments. A retail store might want better stock control. Once the problem is defined, it becomes easier to pick the right tools and track results. Ask yourself what is slowing down your progress today and target that first.

2. Build on reliable and clean data

AI decisions depend on the quality of data. Bad data will give bad results. Kirsten Poon has seen projects fail because the data was missing, incorrect, or outdated. She advises checking your data sources before building any AI system. This means fixing missing entries, removing duplicates, and confirming accuracy. In one project, Kirsten Poon helped a manufacturing company replace a faulty logging system that recorded wrong temperature readings. Once fixed, the AI predictions improved instantly. Clean data helps AI make better and more consistent decisions.

3. Keep projects small at first

Large AI projects can overwhelm teams and delay results. Kirsten Poon suggests starting small. A factory might first use AI to predict maintenance needs for one machine rather than the entire production line. This smaller scope gives faster results and helps teams learn how to work with AI. Once the first project is successful, you can expand to more areas. Starting small builds momentum and reduces risk.

4. Involve the right people early

AI projects work best when the right people are involved from the start. Kirsten Poon says this includes not only data scientists but also the people who understand daily operations. She recalls a warehouse project where the AI system struggled because it did not reflect how staff handled stock. When warehouse employees joined the discussions, they suggested small changes that made the system much more accurate. By involving frontline workers early, companies create AI tools that match real-world needs.

5. Measure results and adjust quickly

AI is not something you set up once and forget. Kirsten Poon advises tracking results and making changes when the AI is not meeting expectations. This starts with clear metrics. If your AI forecasts product demand, measure how accurate it is and how it affects sales. If accuracy drops, find out whether the model needs retraining or if market conditions have changed. Acting quickly keeps AI useful and relevant.

Putting the lessons into action

These five lessons work best together. Start with a problem that matters. Use clean and reliable data. Begin with a small project. Involve people who understand the work. Monitor results and adjust as needed.

Kirsten Poon has seen how AI can help people work smarter. In one project, she helped a packaging company predict which machines would need service. This cut downtime by 30 percent and allowed technicians to focus on urgent repairs. The AI did not replace jobs; it made them more effective.

Kirsten Poon says that companies do not need the most advanced AI tools to see results. What matters most is clarity, good data, and a process that fits how people work. Small, focused changes often bring more value than large, costly systems.

If you are starting your first AI project, pick one goal and track progress closely. If you already use AI, check whether it is still solving the problems you set out to fix. Bring together technical experts and the people who know your processes best, and make sure they share insights often.

The success of AI depends on how well it serves the real needs of the business. These five lessons from Kirsten Poon are a simple guide for any company, from small local shops to large manufacturers.

Kirsten Poon has seen both strong and weak AI projects. The difference, she says, comes from focus and discipline. Companies that follow a clear plan, involve their people, and monitor results see lasting value. Companies that rush without a plan often have to start over.