( Independent Practice / 2026 )
Lighthouse
Designed a three-product ecosystem end to end as the sole designer: a mobile marketplace and planning app, a professional web platform, and a dual-sided enterprise SaaS portal, spanning 130 or more portal screens plus complete mobile and web experiences
- My role
- Freelance UI/UX Designer
- Timeline
- 2026
- Category
- ConTech / AEC Marketplace / B2B SaaS
- Status
- New design, delivered
Tools Figma
Generated schematic, not a screenshot. Real imagery to follow.
( The account )
Context
Lighthouse is a construction-tech startup building a unified digital platform for India's architecture, interior design, and construction material sourcing ecosystem. It sets out to connect three groups that already do business with each other but have no shared system to do it in: manufacturers (brands), authorized distributors (resellers), and design professionals (architects, interior designers, contractors).
The ambition was never a single app. Serving all three groups properly meant three products with genuinely different jobs. Design professionals needed a mobile-first experience for discovery, material planning, and procurement on the move. The same professionals needed a desktop counterpart for the heavier work of browsing catalogs, managing quotes, and maintaining a public professional presence. Brands and resellers needed something else entirely: an enterprise SaaS portal for catalog management, reseller authorization, inventory, and quote response.
I joined with a proposition and no product. There was no existing design, no component library, and no established interface language on any of the three surfaces. The scope was end to end across all of them, which came to 130 or more screens on the business portal alone plus complete mobile and web app experiences, delivered as the sole designer.
The problem
India's construction and building materials industry is among the largest in the world, and almost none of it runs on software. The supply chain is fragmented, manual, and opaque at every level, and the specific failures differ by who you ask.
For design professionals, sourcing is relationship-based and physical. Finding materials means visiting showrooms, making calls, and comparing prices across disconnected channels with no single catalog spanning tiles, paints, cement, flooring, plumbing, electricals, and lighting. Pricing varies widely between suppliers and real-time stock is rarely visible, so buyers check with several sellers before committing anything. Planning materials for a multi-room, multi-phase project demands spreadsheets and domain expertise that less experienced professionals do not yet have. And unlike tech or creative industries with LinkedIn and Behance, AEC professionals have no dedicated platform to show completed work, build a verified reputation, or be discovered on merit rather than word of mouth.
For resellers, the operational load is the problem. Managing inventory across multiple brands means juggling different catalogs, pricing structures, variants, and availability statuses without a unified tool. Getting authorized by a new brand involves back-and-forth email, physical document submission, and long waits with no visibility into approval status. Quote requests arrive through WhatsApp and phone calls, competing quotes are invisible, SLA windows are tight, and there is no analytics layer to tell a reseller their win rate or how competitive their pricing actually is.
For brands, distribution is paper-heavy. Authorizing resellers has no centralized system for verifying credentials, tracking authorization across regions, or managing compliance documents such as GSTIN, PAN, and business registration certificates. Catalog management runs on spreadsheets or legacy systems with no bulk upload, no variant handling for size, grade, finish, and color, and no visibility into product performance across the reseller network.
Underneath all three sits the same structural failure. The quotation lifecycle, from requesting a quote through comparing responses, tracking the order, verifying delivery, and handling disputes, is scattered across WhatsApp messages, phone calls, and spreadsheets. Nothing tracks an RFQ through its stages, so quotes get lost, follow-ups get missed, responses arrive late, and delivery disputes have no record to settle against.
The constraint on all of it was breadth. Three products, three user groups, three interface contexts, all built from scratch by one designer.
What I did
The organising principle was that the RFQ is the spine of the whole ecosystem. Every other feature, discovery, planning, catalog, profiles, authorization, inventory, exists to feed a quote request or to answer one. Designing the RFQ lifecycle first and building each product around its own side of it is what kept three separate applications behaving as one system rather than three products sharing a brand.
That lifecycle runs Active, waiting for quotation, then In Progress, seller selected and order confirmed with price, delivery time, and dealer details, then Verification, where the seller uploads delivery proof and the buyer accepts or reports an issue for damage, incorrect product, incomplete order, or quality, then Completed, with a downloadable PDF invoice. Buyers see competing quotes side by side with pricing, delivery time, and validity. Resellers see the same request from the other side, with an SLA countdown, competitive context showing how many others are quoting, and a structured submission form covering unit pricing, delivery timeline, additional charges, and notes.
The **mobile app** serves the professional on site. An AI-powered material planner works as a guided conversation rather than a form, taking space type, budget, and style preference, accepting floor plan uploads, and returning a complete material requirement list with brand-specific suggestions. A curated marketplace organises products by category with transparent starting-from pricing in rupees and stock indicators. A material board holds materials across multiple projects and phases with cost summaries. A social layer carries verified professional profiles, project posts, and short-form Bytes content. Resellers get their own mobile view for active RFQs, brand authorization status, product analytics, and inventory. All of it works in dark and light mode, because the same person uses it on a construction site and at a desk.
The **web app** is the desktop counterpart for the same professional audience. It pairs a content-rich feed of posts, project showcases, and Bytes with a full marketplace browsable by category, application, and brand. Product pages carry variants, installation availability, delivery timelines, specifications, reviews, and a nearby-reseller map. Professional profiles hold tabbed portfolios and project metadata including certifications such as LEED and GRIHA. A persistent right sidebar keeps Active RFQs, a Resellers Near Me map, and global search on every screen, so procurement context stays visible rather than living in a section users have to navigate to.
The **business portal** serves the supply side and is the largest of the three at 130 or more screens. Both brands and resellers enter through one structured four-step registration covering personal and business information, professional overview, legal information, and review, collecting GSTIN and VAT certificates, PAN cards, business registration, authorized signatory IDs, and brand authorization letters in a single session. A pre-registration checklist front-loads document gathering, and applicants track approval through a visible timeline. Past onboarding, the portal splits: brands manage catalogs with SKU, category, MRP, reseller count, publish status, and bulk upload, plus a reseller network module handling regional authorization, categories, documents, and activity history. Resellers manage brand authorization requests with clear approved, pending, and rejected states, inventory with variant-level control, and the RFQ queue filtered by New, Near SLA, Quoted, Won, Rejected, and Expired.
The portal runs on a dark-theme visual language distinct from the light consumer surfaces, and the whole platform is built for Indian business workflows natively: rupee currency, GSTIN and PAN compliance, city and state level authorization, and local entity types including Private Ltd, LLP, Partnership, and Proprietorship.
Outcomes
- Designed a three-product ecosystem end to end as the sole designer: a mobile marketplace and planning app, a professional web platform, and a dual-sided enterprise SaaS portal, spanning 130 or more portal screens plus complete mobile and web experiences
- Built three design systems from scratch, covering a dark and light mode mobile component library, a light-themed web system, and a bespoke dark-theme portal system for data-heavy administrative work
- Digitized the full RFQ lifecycle from quote request through competitive bidding, seller selection, delivery verification with issue reporting, and invoicing, replacing a process previously run across WhatsApp, phone calls, and spreadsheets
- Designed supply-side tooling supporting catalog management at 234 or more products and reseller network management at 45 or more active resellers from a single brand interface
- Introduced AI-guided material planning to a category that had relied entirely on manual research and word of mouth, compressing sourcing from days of showroom visits and phone calls to a guided session
- Created a professional trust layer for the AEC industry, with verified profiles, project portfolios carrying certifications such as LEED and GRIHA, ratings, and merit-based discovery, in a market where reputation previously travelled only by referral
- Built the platform for Indian business workflows natively, covering rupee pricing, GSTIN and PAN compliance, city and state level authorization, and local entity types
What I learned
Designing for three user groups taught me to look for the shared failure before designing the shared system. My first instinct was to treat buyers, resellers, and brands as three products with three problem statements, and the scope that implied was not survivable for one designer. Recognising that a buyer losing a quote in WhatsApp and a reseller missing an SLA deadline are the same broken transaction seen from opposite ends collapsed three feature sets into one lifecycle with three views. I now look for that shared failure first on any multi-sided product, because it is what decides whether the scope is one problem or several.
The instinct toward a single design system is not always right. Consistency is usually the goal, and with a solo designer and three surfaces the case for one system was strong on effort alone. But a reseller doing hours of catalog administration and an architect browsing materials for ten minutes have genuinely different needs from an interface, and one system serving both would have been a compromise neither side asked for. Deciding by user context rather than by maintenance cost was the right call, and building three narrower systems turned out more tractable than one system general enough to cover everything.
Compliance-heavy onboarding is an information design problem before it is a form design problem. I initially thought about registration as a flow to optimise, when the actual failure was users starting without the documents the flow would eventually demand. Moving the requirement in front of the flow rather than distributing it through the flow was a small structural change that addressed the real cause. It shifted how I think about drop-off generally: the fix often sits before the step where the drop-off shows up.
( Tags )
- ConTech
- AEC
- Construction Technology
- B2B Marketplace
- B2B SaaS
- Enterprise SaaS
- Multi-Sided Platform
- Mobile App Design
- iOS
- Android
- Web Application
- Design Systems
- Dark Mode
- Light Mode
- AI Product Design
- AI Assistant
- Conversational UI
- Procurement
- RFQ Management
- Supply Chain
- Inventory Management
- Catalog Management
- Onboarding Design
- KYC and Compliance UX
- Professional Networking
- Social Feed
- Marketplace Design
- Information Architecture
- Interaction Design
- End-to-End Design
- Solo Designer
- Freelance
- India
- Figma