Our work, in businesses you can open and judge for yourself

The problem

Selling fragrance online is hard for an obvious reason: nobody can smell it. Everything has to be carried by product imagery, description and trust signals. On top of that, the Indian market expects cash on delivery, which brings its own fraud and return problems.

What was built

• Shopify storefront with a custom theme, tuned for product-page speed
• A repeatable AI product-imagery pipeline so every bottle shot matches the same brand system — same lighting, same treatment, no drift between batches
• Cash on delivery with OTP verification, to cut fake orders
• WhatsApp as the primary customer channel, since that is where Indian buyers actually reply
• Product and collection schema so the catalogue is readable in search

Where it stands

The store is live and trading. Because it is ours, we can show the whole thing — the theme code, the imagery workflow, the checkout decisions — rather than describing a project you have to take on trust.

The problem

Independent hotels are badly served by the software market. Enterprise platforms are priced for chains, and the affordable Indian systems were not built around AI. Meanwhile the front desk loses hours a day to WhatsApp enquiries, booking confirmations and review replies.

What was built

• A property management system covering bookings, rooms, rates and reporting
• WhatsApp handling for guest enquiries and booking confirmations
• Automated review responses, which affect Google Maps visibility directly
• Every feature used in daily operations before it is considered finished

Where it stands

The hotel operates on it. That is the point: when we tell a hotelier we understand their operation, it is because we run one. It is also why Roomline exists — the product came out of the problem, not the other way round.

The problem

Indian government procurement runs across GeM, CPPP and a long list of state portals, each with its own interface. For a small supplier, simply finding the relevant tenders is a daily manual job before any bidding work begins.

What was built

• Aggregation and normalisation of listings from multiple portals
• Search and filtering by category, value and eligibility
• Structured so an AI matching and eligibility-screening layer can sit on top

Where it stands

It does the job it was built for: the tenders worth bidding on surface without anyone trawling portals to find them. We are actively widening coverage toward every state portal in India, and building the AI matching and eligibility screening that turns a list of tenders into a shortlist of winnable ones.

The problem

A new fragrance brand has no history to lean on. Product imagery, naming and page copy have to establish the brand on their own, and the store has to sell from the first week rather than after a long ramp.

What was built

• Shopify storefront on the same theme system as Attardaan India, re-skinned for a different brand voice
• A locked visual system for bottles, posters and banners so every product image matches
• Product and collection schema so the catalogue is readable in search from day one
• WhatsApp for pre-sales questions and order support

Where it stands

Live and trading. Launching a second brand on the same system is the clearest test of whether the system is reusable — this is what a client store gets: the same storefront, imagery and messaging stack, built for their brand.

The problem

Electronics is unforgiving. Buyers arrive knowing exactly what they need and filter hard on specification, compatibility and price. Get the data model wrong and the store either buries the right product or sells the wrong one. On top of that, margins are thin enough that fulfilment and returns handling decide whether a line is worth stocking.

What was built

• A catalogue structured around real specification data rather than free text
• Filtering and comparison built for buyers who know what they want
• Product schema so listings surface for specific part and spec searches
• Cash on delivery with OTP verification alongside UPI and cards
• WhatsApp for pre-sales questions, which is where most component queries start

Where it stands

Trading as our own business. It is the counterpart to Attardaan India: same underlying system, a completely different buyer. One is bought on feeling with no way to sample the product; the other is bought on a datasheet. Building both taught us where an ecommerce system has to flex.

The problem

A destination venue sells something a hotel does not: a single, high-value, once-only booking made months ahead, usually by someone who has never seen the place. Every enquiry is worth chasing properly, the sales cycle runs long, and the decision is emotional before it is practical. Meanwhile a long-horizon land development has to be marketed alongside it without the two confusing each other.

What was built

• Positioning and landing pages built for high-intent wedding enquiries
• Enquiry capture and follow-up over WhatsApp, where these conversations actually happen
• Our property management system running the venue side
• Paid campaigns with session recording and heatmap review feeding page changes
• A structure that lets the development story grow without diluting the venue offer

Where it stands

Operating, and expanding. It is the largest thing we run, and the clearest test of whether our hospitality systems hold up outside a conventional hotel — a venue, a resort and a development sharing one set of tools.