Dishtective

Inspiration

The elderly make up the majority of hawkers. Since they tend to not be familiar with technology, they might face difficulties in setting up their “digital restaurant” on food delivery apps which are on the rise post-pandemic due to its convenience for patrons.

What it does

Classify the dish into categories, then generate a suggested description for the hawker to use. This process would be a human-over-the-loop AI ML algorithm.

How we built it

Challenges we ran into

The model have difficulties in recognising noodle-like food, e.g. spaghetti. we changed the model and the situation improve, albeit not the best.

Accomplishments that we're proud of

We integrated openAI API into our application and the output for the prompt for suitable suggestions of menu descriptions is really relevant and suitable!

What we learned

-Train image recognition machine learning model -Work with openAI API in Python

What's next for Dishtective

Help more hawker/SME owner in F&B Industry out there digitise their menu and integrate their service with food delivery app easily. We hope Dishtective can help in Singapore's effort in progressing towards a smart digitised nation in the future.

Built With

• python