Built. Shipped. Running.

Furniture Concepts 2.0 needed to rebrand into Cohuman—a modern, high-end office furniture manufacturer for architects, interior designers, and corporate offices. Their previous site was structured like a consumer e-commerce store, creating friction for corporate B2B buyers who order in bulk and require custom specification quotes rather than instant checkout.

Rebuilt the entire digital architecture into a high-converting digital showroom & specifier engine. Created a Next.js 16 catalog repository, 3 specialized lead capture pipelines (Product Enquiry, B2B Bulk Order, Showroom Visit), pre-filled quote dialog context, Zod API validation, Resend email routing, and sub-second mobile page performance.

A premium furniture manufacturer with a strong Instagram presence (13K+ followers) and a physical showroom in Surat — but no digital storefront. All traffic had nowhere to convert. Customers would DM asking for catalogs, and the team would manually send PDFs. A lot of interest, zero scalable infrastructure.

A premium, editorial-style website with immersive 3D product interactions, a full browsable catalog with category filters, WhatsApp enquiry on every product, and a high-performance Next.js build optimized for mobile. The design elevated their brand from a social-media-only presence to a proper digital storefront that matches the premium nature of their products.

Welcome Palace runs three distinct businesses under one roof in Piplod, Surat — 15 luxury rooms, a private banquet hall for 125–175 guests, and in-house Chandni Chowk Chaat & Catering. All of it lived behind one generic enquiry form, with no way for a bride booking a wedding hall to find the same urgency as a couple booking a honeymoon suite.

A fast, local-SEO-optimized Next.js site with dedicated journeys for rooms, the banquet hall, the "Shadi Wala Ghar" family wedding-stay offer, and catering — each card routing straight into a pre-filled WhatsApp enquiry so every space converts on its own terms.

A furniture retailer with 3 showrooms across Gujarat had zero digital memory. Staff couldn't recognize returning customers. Leads from Instagram, Google, and walk-ins leaked into separate WhatsApp chats with no tracking. There was no sales pipeline, no performance data, and no way for the owner to see what was happening across locations without calling each manager.

Consent-based face recognition at showroom entries that identifies returning customers and alerts staff on WhatsApp. A custom CRM with omnichannel lead capture (Instagram, Facebook, Google, walk-in). An AI WhatsApp assistant that speaks Hindi, Gujarati, and Hinglish — grounded on the live catalog with real pricing. Automatic salesperson assignment and a real-time sales pipeline dashboard.

The quote-to-cash and production layer. Salespeople generate GST-compliant quotations that land in the customer's WhatsApp as PDFs with one-tap approval. Approved quotes become orders on a pipeline kanban. Payment schedules track advances and dues with automated WhatsApp reminders. On the factory side, every order item is assigned to a workshop and tracked stage-by-stage through a bilingual Gujarati–English PWA — with delay watchdogs and customer notifications the moment production starts, completes, or is ready for dispatch.

• ▪M1 — Showroom Intelligence (Face Recognition)
• ▪M2 — Customer CRM & Lead Capture
• ▪M3 — AI WhatsApp Assistant
• ▪M4 — Sales Pipeline & Dashboard
• ▪M5 — Salesperson Assignment Engine

• ▪2A — GST Quotations, PDF Documents & One-Tap Approval
• ▪2A — Orders, Pipeline Kanban & Payment Tracking
• ▪2B — Workshop & Production Stage Engine
• ▪2B — Bilingual Workshop PWA (Gujarati + English)
• ▪2B — Production Board, Notifications & Delay Watchdog

Every year, urban Indians collect blood test after blood test, but the reports sit as unread PDFs. Nobody tracks what changed since last time, generic AI summarizers overstep into diagnosis, and doctors still receive raw numbers instead of a clear story of what's trending and what needs attention.

A private, doctor-reviewed personal health graph. Users upload a lab report and get source-linked biomarkers explained in plain language, tracked across a timeline, with retest reminders and an optional doctor review workflow. Every explanation ships under strict medical-safety guardrails — the AI explains and organizes, doctors diagnose and prescribe, never the reverse.

• ▪Report Upload → Private Storage → Processing Job
• ▪Schema-First Biomarker Extraction
• ▪AI-Assisted, Safety-Gated Explanation
• ▪Health Timeline & Retest Reminders
• ▪Optional Doctor Review Workflow