AnalyzeUp

• 

  landing page

• 

  landing page

• 

  landing page

• 

  Dashboard page

• 

  Predict the strategy as per the data injected and the status of current data it gives proper detailed strategy for it

• 

  It gives predictions for best possible restocks to be made using AI

• 

  The app has change of theme feature as well Light and Dark

• 

• 

  Here detailed graph of all data and profits made can be reviewed since i have made only 2 orders for demo so it shows one line graph

• 

• 

• 

  We can import csv files and it will populate the entire app with its data

• 

  Supplier info can be added

• 

• 

• 

• 

• 

Inspiration

Small and medium businesses in India often rely on spreadsheets or fragmented tools to manage inventory, orders, and suppliers. We wanted to build a unified, intelligent platform that brings real-time data visibility and AI-generated business strategy to any SMB — without needing a dedicated data analyst.

What it does

AnalyzeUp is a multi-tenant inventory management and business intelligence dashboard. It lets businesses:

• Track products, stock levels, categories, and SKUs in real time
• Manage suppliers and raise purchase orders
• Record and visualize sales and purchase transactions via dynamic charts
• Get AI-powered low-stock alerts — AI analyzes average daily sales and supplier lead time to predict stockout risk and recommend reorder quantities
• Generate a full AI business strategy — summarized recommendations, risk factors, and expected outcomes based on live sales and product data
• Export reports and visualize data with an interactive data visualizer

How we built it

• Frontend: Next.js 15 (App Router, Turbopack), React 18, TypeScript, Tailwind CSS, shadcn/ui (Radix UI primitives), Recharts for data visualization, Framer Motion for animations
• AI Layer: OpenAI — two server-side AI flows: AI StockAdvisorFlow for per-product reorder intelligence and Business StrategyFlow for holistic growth strategy generation. All AI I/O is validated with Zod schemas.
• Backend / Database: Firebase Firestore (real-time NoSQL), Firebase Authentication, multi-tenant architecture with tenantId scoping across all collections
• Forms & Validation: React Hook Form + Zod for type-safe, validated forms throughout the app
• PDF Export: jsPDF + html2canvas for report generation
• CSV Parsing: PapaParse for data import flows

Challenges we ran into

• Building a robust multi-tenant data architecture in Firestore — every entity (Product, Supplier, Order, Transaction) is scoped by tenantId, ensuring complete data isolation between businesses
• Structuring Genkit AI flows with strict Zod input/output schemas so Gemini always returns structured, parseable JSON rather than free-form text
• Managing real-time Firestore listeners efficiently in React without stale data or memory leaks, using custom hooks (use-collection, use-doc)
• Handling Firebase initialization safely in Next.js SSR/client boundaries using a ClientOnly wrapper and provider pattern

Accomplishments that we're proud of

• Two fully functional, production-ready AI flows using Google Genkit and Gemini 1.5 Flash with Zod-validated structured output
• A complete multi-tenant SaaS architecture on Firebase — sign up, and your data is fully isolated from other tenants
• End-to-end type safety from Firestore documents → Zod schemas → React components → AI prompts
• A polished, responsive dashboard UI with dark mode support, animated hero section, and smooth UX built with shadcn/ui + Framer Motion

What we learned

• How to architect Genkit AI flows with structured prompts and Zod schema enforcement for reliable AI output in production
• Deep understanding of multi-tenant Firestore design patterns — collection scoping, security rules considerations, and tenant-aware queries
• How to integrate Next.js 15 App Router server actions with Genkit flows ('use server' AI functions called from client components)
• Managing complex real-time state across a large dashboard without global state libraries by using React Context + custom Firebase hooks

What's next for AnalyzeUp

• Role-based access control (admin, staff, viewer) within each tenant
• Demand forecasting with historical transaction data using more advanced Gemini prompts
• WhatsApp / email alerts for low-stock notifications
• Integration with Indian GST billing and invoice generation
• Mobile app via React Native with the same Firebase backend

Built With

• date-fns
• firebase-(firestore-+-authentication)
• framer-motion
• gemini-1.5-flash
• google-genkit
• html2canvas
• jspdf
• lucide
• next.js-15
• openai
• papaparse
• radix-ui
• react-18
• react-hook-form
• recharts
• shadcn/ui
• tailwind-css
• typescript
• zod