Where are we going for lunch?

I eat out more than I should and I am not proud of this fact. I’ve tried to develop mechanisms like setting $15 budgets per day but with the inflation these days…. lets just say my belly wins the hunger games and I end up going out of budget. I’m sure I’m not alone.

The crux of this post is not about healthy eating habits or cooking but its really about capitalizing on this habit that millions have. When you eat out often, you never have a taste for anything so trying to decide what to eat or what restaurant to go to becomes stressful.

This is when the idea bulb lit up! How about creating an app or service that helps you decide what to eat, based on what you don’t want to eat? The hard part is knowing what you want to eat but knowing what you don’t want is easy.

I quickly realized this was a machine learning problem. I needed good data to create a meal prediction model by taking in ingredients as inputs. I would need to send out a questionnaire to friends (preferably ones that eat out a lot) to learn more about how they decide what to eat. I could even take it as far as creating a database of the menus of nearby restaurants so the model can suggest a specific meal at a restaurant. However, this adds a thick layer of complexity that increases the magnitude of the problem because restaurants constantly refresh their menus, prices change and it will take considerable effort to compile all that data.

I’ll let this idea marinate some more but until next time, stay tuned.