What's for Dinner/Aaj Kya Banega? A 20B model trained to solve my Mom's kitchen crisis without the chatty fluff
This is a submission for the Hacktoberfest Weekend Challenge: Build for a Friend What I Built The most dangerous question in any Indian household: "Aaj khane mein kya banau?" (What's for dinner?) Obviously, most days I'll say, "Make whatever you want." And there's a good chance either Mom gets frustrated with that answer, or we're having the same dish for the fourth time that week. So I built…
In India, the age-old question "Aaj khane mein kya banau?" or "What's for dinner?" often plagues households. To tackle this issue and help his mother, the author built a meal-planning assistant called "What's for Dinner?" An AI-powered application that suggests up to three Indian home-cooked dishes based on the actual ingredients available in the fridge.
It also remembers what the family has eaten recently and avoids repeating those suggestions. The project utilizes a 20B open-weight model called "gpt-oss-20b," fine-tuned on a synthetic dataset of Indian recipes, Hinglish conversations, and local ingredient names. Tinker API was used to sample LLM-generated suggestions while Backboard API maintained a history of recently selected meals to prevent repetition.
The entire application is deployed as a single Web Service on Render, making deployment simple and cost-efficient.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
