Agro-Assist AI

• 

  this is login page and for now any username and password you can get access of the app

• 

  this is our home page when you will get all info about daily market and recent scan you have done

Inspiration

Growing up in Rwanda, I often saw farmers struggling with plant diseases and having no immediate access to agronomists. Crops were lost, livelihoods were affected, and solutions were often expensive or unavailable locally. I wanted to create a tool that empowers farmers directly, giving them expert guidance in real time through the devices they already own.

What it does

AgroAssist AI transforms a smartphone into a personal agronomist:

AI-Driven Visual Diagnostics: Detects diseases like Late Blight in potatoes and Maize Lethal Necrosis, highlights affected areas, and explains the reasoning behind every diagnosis.

Recovery Tracking: Tracks plant health over time with chronological scans and trend graphs.

Localized Treatment Guidance: Recommends RAB-approved pesticides/fungicides with exact dosages and provides organic remedies.

Accessible Interface: Offers text-to-speech, live AI voice Q&A, and high-contrast UI for outdoor use.

Market & Weather Insights: Shows real-time crop prices, “Safe to Spray” alerts, and tools to create marketing content.

How we built it

Developed as a Progressive Web App (PWA) with offline caching.

Integrated Gemini 3 Flash for vision-based diagnostics and reasoning.

Added Gemini 2.5 TTS and Native Audio for voice-guided interaction.

Connected real-time market and weather data using Google Search and Maps.

Designed a high-contrast, multi-language interface optimized for outdoor use and low literacy levels.

Challenges we ran into

Collecting diverse crop images to train the AI.

Ensuring the app works offline in rural areas with low connectivity.

Integrating multiple AI models smoothly in real time.

Localizing recommendations to Rwandan-approved treatments and languages.

Accomplishments that we're proud of

Developed a tool that acts like a field agronomist in your pocket.

Enabled farmers to track recovery progress over time.

Created a system that bridges technology and local agriculture, making AI accessible and actionable for smallholder farmers.

What we learned

Technology is most impactful when it is practical, inclusive, and culturally relevant.

Multi-modal AI can be integrated into a single user-friendly interface without overwhelming the user.

Clear communication of AI reasoning builds trust and adoption among non-expert users.

What's next for AgroAssist AI

Expand coverage to more crops and regional plant diseases.

Add community-driven knowledge sharing between farmers.

Integrate predictive analytics for crop yield forecasting.

Improve real-time alerts and personalized treatment suggestions using AI learning from historical scans.

Built With

• browsergeolocationapi
• custompersistancelayer
• fontawesome6
• gemini2.5flashnativeautio
• gemini2.5flashpreviewtss
• gemini3flash
• gemini3pro
• gemini3proimage
• googlesearchtooling
• lucide
• node.js
• react19
• serviceworkers(pwa)
• svgoverlaysystem
• tailwindcss
• telemetrysystem
• veo3.1fast
• webaudioapi