PlantLens Pro: Turning Local AI into a B2B Side Hustle (Directly targets the "Side Hustle" theme)

A developer built PlantLens Pro, a local-first, offline desktop app for independent plant nurseries that uses LLaVA 1.5 7B via Ollama to diagnose plant diseases from a webcam photo and generate a branded, printable "Care Recipe" PDF with in-store product recommendations. The app pairs a Next.js/React/TypeScript frontend with a Tauri v2 desktop wrapper, and the developer argues local inference enables a $50 one-time license by eliminating API costs, cloud timeouts in poor-connectivity greenhouses, and privacy concerns over uploading shop data.

This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle https://dev.to/challenges/hacktoberfest-week2-2026-10-12 PlantLens Pro is a local-first, offline B2B desktop application designed specifically for independent plant nurseries and boutique garden shops. When a customer brings in a sick plant, the shop owner uses their counter webcam to snap a photo within the app. A local open-weight vision model diagnoses the disease and instantly generates a shop-branded, printable "Care Recipe" PDF for the customer. The Business Case: Nursery staff waste hours every day answering the exact same question: "Why are my plant's leaves turning yellow?" PlantLens Pro solves this tedious bottleneck. It automates the consultation, provides the customer with a premium takeaway, and includes a built-in upsell section recommending specific fertilizers or soils available in the store. Because it runs locally with zero API costs, it is sold to small shops as a highly attractive $50 lifetime license, rather than a difficult-to-sell monthly SaaS subscription. https://plantlens-ai-pro.vercel.app/ https://plantlens-ai-pro.vercel.app/ https://github.com/r67251491-coder/PlantLens-Ai-Pro.git https://github.com/r67251491-coder/PlantLens-Ai-Pro.git The project is built around running open-source vision models completely locally. The AI Core: I used LLaVA 1.5 7B via Ollama running locally on port 11434. I engineered a strict system prompt instructing the model to act as a botanical diagnostician and output pure JSON containing the plant name, issue, action steps, and an in-store product recommendation. The Frontend: Built with Next.js, React, TypeScript, and Tailwind CSS. The UI features a high-contrast, dark-mode professional dashboard designed for efficient counter use, complete with hardware status polling to ensure Ollama is running and keyboard shortcuts for fast operation. The Desktop Wrapper: I used Tauri v2 to package the Next.js web build into a lightweight, native desktop executable that easily interfaces with the local webcam and printer. Building this with a closed-source, cloud-based API would have completely destroyed the business model for three reasons: The "Greenhouse Wi-Fi" Problem: Garden centers, rural nurseries, and metal-framed greenhouses notoriously have terrible or non-existent internet connections. Cloud APIs time out. By running LLaVA locally via Ollama, PlantLens Pro works flawlessly completely offline. Zero Marginal Cost Pricing: Small retail shops operate on razor-thin margins and hate recurring $30/month software subscriptions. Because my AI inference cost is literally $0, I can sell this as a flat one-time fee. It becomes an immediate, no-brainer impulse buy for a stressed shop owner. Local Privacy: Boutique shops don't want to upload their proprietary inventory issues, supplier tags, or customer interactions to a corporate server. Everything stays on their own hard drive. Best Local Inference Build Best Business Utility Thanks