Background

In the past year, I have released two children's books, started a YouTube channel, and deployed two games on the Play Store, and I am currently working on building a full-scale app with the help of my sidekick, Generative AI. Three years ago, Generative AI changed how we do everyday tasks. Many once time-consuming tasks are now far more manageable thanks to Generative AI. If you haven't tried tools like ChatGPT, I strongly recommend doing so today. I have experimented with these generative AI tools, and I now use them regularly.

About seven years ago, I stopped writing code and focused solely on increasing my product management skills. However, with the help of Generative AI, I have started writing code again. While AI handles the heavy lifting, my experience helps fine-tune and debug when needed. Still, my experience is helping me fix a few bugs generated by the AI agent. In a future post, I will break down exactly what platforms I use for specific tasks.

With the help of Generative AI, you can accomplish many tasks; the possibilities are endless and left to your imagination. As I study for my MSc in AI for Business, I have come to realize the true power of these tools. Here are three main ways I utilize Generative AI as a product manager.

1. Brainstorming

When my child was born, I realized the market for children's books was about $11 billion, so I started brainstorming what parents buy for their kids from various parts of the world. With the help of Generative AI, it became evident that children's books are a promising market to venture into, and there are also options for digital distribution. One of the best parts was Gemini gave me step-by-step instructions on how to publish a book on Amazon. Using generative AI, I was able to brainstorm the types of books I could create, and I landed on graphical books with simple words. Of course, I had always been aware of Kindle Self-Publishing, but I was unable to produce books due to a lack of artistic skills. Now, with the help of Canva and Gemini, I was able not only to create a book skeleton but also to publish it on Amazon.

2. Data Analysis and Visualization

As a product manager, I always make decisions based on data. However, two years ago, I had to rely on building panda visualizations or use tools like Power BI to create good visualizations. Now, I can pass the dataset through a Generative AI model and ask it to build graphs for me. Additionally, Generative AI can also summarize large datasets using plain simple English. You can even ask simple questions and get responses based on the dataset you provided. This simple summarizing feature can help you save a significant amount of time. However, remember that Generative AI can sometimes produce incorrect or hallucinated results, so always double-check the responses before making any critical decisions.

3. Prototyping

One of my favourite things I use Generative AI for is building prototypes. AI Agents like Copilot and Cursor can write thousands of lines of code and even help you fix any bugs you are facing in your code. My productivity as a product manager has increased drastically with the help of AI Agents. With the help of Data Analysis from Generative AI, deciding on what prototype to build becomes easy. Here is a simple example: over the past few weeks, my child has been attending a private daycare centre. With the help of Copilot, I was able to build a quick prototype to allow my babysitter to track our child’s activity. With the help of GitHub, I was able to host it and pass it over to the babysitter. The concept of writing code using AI Agents is Vibe Coding; in a future post, I will cover the pros and cons of Vibe Coding.

Conclusion

As a product manager, I strongly believe in data-driven decision-making and a fail-fast methodology. Generative AI has helped me brainstorm ideas, perform data analysis, extract data from trustworthy sources, and build quick prototypes to share with friends and family, allowing me to understand if the product I’m building is worth investing time in. All these things wouldn’t have been possible without some trial and error, as well as prompt engineering.

Feel free to connect with me on LinkedIn or visit my site here.