Inspiration
Farmers and poultry owners often lack fast access to expert diagnosis. AgroShield AI was inspired by the need for affordable, instant agricultural health assessment using only a smartphone.
What it does
AgroShield AI analyzes plant leaves and poultry images to detect diseases in seconds. Using Gemini 3, it provides symptoms, treatments, prevention tips, and verified expert resources.
How we built it
We built the frontend with React + TypeScript and Tailwind CSS. Images are sent to Gemini 3 Flash for multimodal analysis. Google Search grounding ensures updated treatments, while results are parsed via structured JSON.
Challenges we ran into
Ensuring accurate diagnoses from varied image quality, designing strict AI prompts, handling real-time camera input, and maintaining low-latency responses were our main technical challenges.
Accomplishments that we're proud of
We delivered a production-ready dual-mode diagnostic system with real-time AI vision, grounded medical guidance, offline history, and a clean mobile-first UI.
What we learned
We gained hands-on experience with Gemini multimodal APIs, prompt engineering, structured outputs, and building AI products for real-world environments.
What's next for AgroShield AI
We plan to add multilingual support, voice guidance, regional crop models, offline inference, farmer dashboards, and IoT sensor integration.
Log in or sign up for Devpost to join the conversation.