Lamp & Glow: Custom Next.js E-commerce & AI

The Core Problem: Running a handmade-lighting export brand across multiple international markets (Pakistan and UAE) was becoming an operational nightmare for a solo founder. Managing multi-currency checkouts, crafting SEO-optimized product descriptions for hundreds of variants, and handling high-volume WhatsApp customer inquiries was bottlenecking growth and operations.

The Technical Solution: I built a complete ecosystem replacing standard Shopify limitations with a highly customized stack.

• Web Storefront (Next.js): Built a high-performance Next.js storefront supporting dual markets (PKR/AED) with dynamic duties/taxes messaging scoped specifically to the UAE.

• AI Content Pipeline: Integrated Groq (llama-3.3-70b) with strict anti-hallucination rules to automatically rewrite and optimize ~92 products and ~190 variants, bypassing manual data entry.

• Internal Operations App (Flutter): Developed a Riverpod 2/go_router Flutter app connected to the Next.js/MongoDB API to manage craftsmen, attendance, wallets, and shift-lookups (handling complex UTC vs PKT timezone logic).

• WhatsApp AI Agent: Deployed a custom WhatsApp sales bot using Groq SDK tool-calling directly over Prisma queries.

Measurable Business Impact: The new architecture enabled the brand to scale to hundreds of monthly orders across two distinct markets without hiring extra staff. The AI pipeline saved over 60+ hours of manual copywriting, while the custom WhatsApp bot successfully auto-resolves over 80% of tier-1 customer inquiries instantly.

Tech Stack: Next.js, Prisma, MongoDB, Flutter, Riverpod, Groq AI, WhatsApp Cloud API.