Project Story: FruitGuard AI - Revolutionizing Berry Care from Diagnosis to Community Inspiration My journey with FruitGuard AI began not in a lab, but from a very personal challenge right here in Leeds, UK. As an international student, I eagerly took on growing strawberries, only to quickly discover how prone they are to various diseases. Faced with ailing plants, I found myself adrift. Diagnosing the specific problem, then figuring out local remedies or where to even buy the right products was incredibly frustrating, especially without easy access to familiar stores or specialized e-commerce sites. This personal struggle, combined with the realization that countless other home growers and small farmers likely face similar hurdles, sparked the idea for FruitGuard AI: an all-in-one, accessible platform to simplify berry plant care.
What it does FruitGuard AI acts as a comprehensive, omnichannel plant doctor and support system for berry enthusiasts.
AI Plant Doctor: At its core, users can simply upload a photo of their affected strawberry, raspberry, blueberry, or cherry plant. Leveraging the power of the Gemini API, our AI instantly diagnoses the disease (e.g., Strawberry Fruit Rot) and provides clear, actionable remedies. This moves beyond mere identification to practical advice.
Integrated E-commerce: Beyond diagnosis, the app intelligently recommends specific products (like specialized fertilizers or treatments) that can help combat the identified disease. Users can filter products by their location (e.g., Leeds, UK) and berry type, view detailed product information, and seamlessly proceed to a shopping cart and checkout. This eliminates the need to scour multiple external sites for solutions.
Community Social Media: FruitGuard AI fosters a vibrant, supportive community. Users can share photos of their plants, post updates on their progress, and engage with other growers to seek advice, share successes, or simply connect. This peer-to-peer knowledge sharing leverages collective experience for everyone's benefit.
Essentially, FruitGuard AI provides a complete cycle of care: diagnose, treat, shop for solutions, and share knowledge, all within one intuitive platform.
How we built it The development of FruitGuard AI was a testament to rapid prototyping and leveraging powerful modern tools, especially considering it was built as a solo project by a first-time coder.
Rapid Prototyping with Bolt.new: The entire application's structure, user interface, and overall development environment were built using Bolt.new. This innovative AI-driven development platform was instrumental. It allowed me to describe my vision in plain language, and Bolt.new generated the foundational code, handling much of the complex setup (like frontend/backend connections, package installations). This dramatically accelerated the development process and made building a full-stack application achievable without prior coding experience.
Gemini API for AI Diagnosis: For the crucial AI diagnosis feature, I integrated the Gemini API. When a user uploads a plant photo, it's sent to the Gemini API, which processes the image using its advanced multimodal capabilities. Gemini then returns a precise disease diagnosis along with contextually relevant remedies. This integration allowed me to incorporate sophisticated AI without needing to train a custom machine learning model from scratch.
Backend & Database: The product listings for the e-commerce section, user posts for the community feature, and diagnostic data are managed through an underlying database system.
Omnichannel Integration: The seamless flow between the AI diagnosis report, the personalized product recommendations, and the social community section was carefully designed and implemented. This required meticulous attention to the user journey to ensure a cohesive and intuitive experience within the Bolt.new framework.
Challenges i ran into Developing FruitGuard AI presented a uniquely challenging, yet rewarding, experience, primarily because I was a solo participant tackling my very first hackathon with no prior coding experience:
Steep Learning Curve for API Integration: One of the most significant hurdles was teaching myself the fundamentals of API integration from scratch. Understanding concepts like HTTP requests, JSON data structures, API keys, and securely connecting my application to the Gemini API required dedicated self-study, countless tutorials, and persistent debugging without immediate expert guidance.
Mastering a New Platform (Bolt.new) on the Fly: Simultaneously, I had to rapidly learn and effectively utilize the intricacies of the Bolt.new platform. While designed for ease of use, navigating its functionalities, understanding how to prompt it effectively, and troubleshooting generated code within an unfamiliar environment demanded continuous exploration and problem-solving under tight deadlines.
Optimizing AI Output for Practicality: Even with the powerful Gemini API, structuring effective prompts and interpreting its responses to extract precise, actionable diagnostic and remedy information for real-world plant diseases was a key challenge. It was a learning process to "speak" the AI's language to get the most useful output.
Integrating Diverse Functionalities as a Solo Developer: Weaving together an AI diagnosis tool, a functional e-commerce platform, and a social media network into a cohesive app is a complex task even for an experienced team. Doing this entirely on my own, while simultaneously learning the underlying technologies, was a significant test of my ability to break down problems and manage scope.
Intense Time Management and Problem-Solving: The extreme time constraints of a hackathon, compounded by the extensive self-learning required for every new concept, demanded ruthless prioritization and an unwavering commitment to overcoming every bug and logical error that arose.
Accomplishments that i am proud of This hackathon provided an unparalleled, intensive learning experience:
The Power of Self-Directed Learning and Resilience: I discovered an immense capacity for teaching myself complex technical concepts (like API fundamentals and platform specifics) under pressure. It taught me the importance of persistence and systematic problem-solving.
Practical Application of AI (Gemini API): I gained invaluable hands-on experience in integrating and effectively prompting a large language model like Gemini to perform specific tasks, moving beyond theoretical understanding to practical application.
Leveraging AI-Powered Development Platforms (Bolt.new): I learned how platforms like Bolt.new democratize app development, enabling individuals without a traditional coding background to bring ambitious ideas to life and dramatically accelerating the prototyping phase. Bolt has truly given me wings as a “no-coder”—I can now successfully deploy my idea without relying on others, even in a short span of time.
Problem Decomposition: Breaking down a large, multifaceted project into smaller, manageable tasks was crucial for making progress, especially while learning new concepts concurrently.
The Value of User-Centric Design: My own frustration as a berry grower directly informed the app's features and flow, reinforcing that solving a genuine user problem is paramount.
Personal Growth and Confidence: Initially, I was underconfident and spent about half of the hackathon (15 days) searching for a teammate without success. That experience forced me to learn on my own and slowly gain confidence. Although my current work may not be of the highest standard yet, I am genuinely happy that I was able to complete it on my own. Balancing this learning with my ongoing exams made it even more challenging, but ultimately it was rewarding to see my idea come to life through practical integration and deployment.
What's next for FruitGuard AI The future of FruitGuard AI is ripe with possibilities, building on the strong foundation established during this hackathon:
Expanding Berry Coverage & Disease Database: I plan to continuously expand the database to include an even wider variety of berry types and more granular classification of diseases, potentially refining Gemini prompts for earlier-stage detection.
Predictive Analytics & Prevention: Leveraging external data sources like local weather information and historical disease patterns, potentially analyzed further by Gemini, to offer proactive alerts and preventative measures to users.
Expert Consultation Integration: Exploring features that would allow users to connect directly with agricultural experts or plant pathologists for more in-depth, personalized advice.
Enhanced Community Features: Adding more interactive social functionalities such as direct messaging between users, the creation of specialized interest groups, and in-app event scheduling.
Global Supplier Onboarding & Localization: Scaling the e-commerce platform by onboarding more local suppliers globally, ensuring users always have access to geographically relevant products and services.
Cross-Platform Refinement: Further optimizing and refining FruitGuard AI for seamless performance across both iOS and Android platforms, ensuring a truly native and performant experience.
FruitGuard AI aims to evolve into the definitive resource for berry growers worldwide, simplifying plant care, fostering a thriving global community, and empowering individuals with intelligent tools.
Built With
- bolt
- gemini
Log in or sign up for Devpost to join the conversation.