ChatGPT supports 16-year-old Aarav Sharma in inspecting and reformatting data as he builds his startup, SmartBin AI . Work that can ordinarily take five or six hours takes about 10 minutes. Aarav integrated OpenAI models into SmartBin AI to turn images of discarded meals into data school cafeterias can use. He started the company after watching a classmate throw away an entire meal. When he asked why so much lunch ended up in the bin, other students said they disliked the dish or had simply been served too much.
SmartBin AI’s camera photographs what remains on a plate and sends the image over Wi-Fi to an analysis pipeline Aarav developed. The model identifies each food and estimates its share of the plate—70 percent pasta, for example—then returns the data to a dashboard. Aarav built the hardware, image-analysis workflow, data pipeline, and dashboard himself. Each month, the system shows food vendors which dishes are discarded most often and where a smaller default serving might help.
At United World College of Southeast Asia in Singapore, the data pointed to unusually high waste from Pad Thai. Aarav recommended cutting the default portion by 20 to 30 percent. The vendor chose a smaller reduction while allowing students to ask for more. Aarav says food waste subsequently fell about 19 percent. The kitchen could prepare less food without limiting students who wanted a larger serving.
Aarav will start Grade 11 at the school in August, but began building years earlier: he was exposed to math at five or six, started 3D printing at nine, then moved to Raspberry Pis—small computers that connect to cameras and sensors. By high school, he was developing SmartBin AI while taking college-level math, physics, and design-technology classes.
Misaligned cameras, underpowered hardware, crashes, and food misclassification forced rebuilds; Aarav still assembles each module, fits the camera and compute unit, and calibrates the software himself. Finding customers was harder. Too young to have an overseas network, he sent cold emails, shared one-page result summaries, and followed referrals. He has since built and deployed SmartBin AI internationally, analyzing more than 20,000 disposals across four continents.
Aarav fits the work around school: a run most mornings, classes, project work at lunch, and time with friends after school, sometimes playing badminton or swimming.
Questions from those classmates are already shaping the next module. After students asked what the device above the bin was doing, Aarav decided to add a touchscreen so they—not only cafeteria managers—can see how much food is being wasted or saved in real time.
