Sow & Grow
Voice-first AI farm companion app
The Challenge
Providing accessible, interactive agricultural guidelines to rural farmers with low internet speeds and varied literacy levels.
Engineering Solution
Designed a voice-friendly Flutter client using local TTS/STT packages. Implemented Gemini Flash 2.0 via a REST service layer (FarmCare) to handle plant image diagnoses, paired with a local-first Flutter Map dashboard showing geospatial plant disease heatmaps.
Key Product Capabilities
FarmHelper AI: Conversational voice assistant featuring Gemini Flash 2.0 chat, image uploads, and Speech-to-Text.
Livestocare Suite: Simple tracker registry for livestock records, vaccination dates, and SMS reminders.
Disease Heatmap: Interactive OpenStreetMap rendering overlay layers of diagnosed crop/animal infections.
Code Implementation
// Dart service invoking Gemini API via Generative AI package
import 'package:google_generative_ai/google_generative_ai.dart';
class FarmHelperService {
final String apiKey = const String.fromEnvironment('GEMINI_API_KEY');
Future<String> askFarmHelper(String prompt, List<int>? imageBytes) async {
final model = GenerativeModel(model: 'gemini-2.0-flash', apiKey: apiKey);
final content = [Content.text(prompt)];
if (imageBytes != null) {
content.add(Content.data('image/jpeg', imageBytes));
}
final response = await model.generateContent(content);
return response.text ?? 'No response received. Please try again.';
}
}Technologies Used
Direct Actions
Architecture Stack
Material 3 UI utilizing Provider state management and geolocator map rendering.
Express service orchestrating user data, regional coordinates, and weather advisory tables.
Performs natural language voice processing and visual diagnosis of crop leaf conditions.
Authenticates accounts, stores uploaded diagnostic logs, and dispatches reminders.