The Ethlore Story: Where D2C Sarees Meet AI

Before we get to the AI part, let me tell you about the saree my grandmother wore to her daughter's wedding in 1974. It was a Banarasi silk, handwoven in Varanasi, with gold zari work that took a master weaver three weeks to complete. It still exists. It will probably outlast most things I own today. That's the product I'm trying to bring to a generation that shops on Instagram and expects next-day delivery.

Ethlore is a D2C handloom saree brand I started with my brother Shuvrangshu, who runs operations full-time from India while I handle strategy, tech, and marketing from the US. We're about 18 months in, doing 10-50 orders per month, and we've invested somewhere between $5,000 and $20,000 depending on how you count the sweat equity. Here's the honest story.

Why Handloom Sarees

The Indian textiles market is enormous and mostly unorganized. Hundreds of thousands of weavers across clusters in Varanasi, Kanchipuram, Dharmavaram, Pochampally, and dozens of smaller towns produce some of the world's most technically complex textiles - and most of them sell through middlemen who extract significant margin while the weavers earn subsistence wages.

The D2C model exists to fix this. By sourcing directly from weavers and weaver cooperatives, we can support weaver livelihoods while making authentic handloom textiles accessible to customers who would otherwise pay retailer markups on the same product running through three levels of distribution. The consumer connects directly with the artisan, the weaver gets better income, and we build a sustainable business on the transparency between them.

That's the vision. The execution is significantly more complicated.

The Sourcing Reality

We source from weaver clusters across multiple regions of India. Each region has different products, different value tiers, different quality tiers, and different lead times. A Banarasi silk takes 3-6 weeks to weave. A Pochampally ikat can be ready in 10 days. A Kanjivaram with heavy temple borders might take two months.

Managing this as a small operation means Shuvrangshu is constantly in motion - visiting weavers, quality-checking samples, managing relationships, handling payments, dealing with the logistics of shipping fragile handwoven textiles across India. He's essentially a full-time operations manager for a startup that's still in its early stages.

The quality variation is the hardest part. Handloom products are handmade, which means inconsistency is inherent. A customer who buys a Banarasi silk and then buys another one six months later might receive something noticeably different - different thread count, slightly different color, different finish. We've had to build quality standards and rejection processes that would feel overkill for a larger operation but are essential for us at this stage.

The Shopify Stack

We run on Shopify. The store is functional but not yet beautiful - we're in the phase where we're learning more from real customers than from optimizing the funnel. The product photography is the bottleneck; handloom sarees are notoriously difficult to photograph well. The texture, the weight, the drape - these are tactile qualities that translate poorly to flat lays and model shots. We're still figuring out the visual language.

The customer experience we've built: product pages with detailed weave descriptions, care instructions, and origin story (which cluster, which technique, sometimes which specific weaver family). Customers who care about provenance are our target audience, and they want that information. Customers who just want a saree and don't care where it came from are probably better served by a marketplace.

We've connected the store to basic analytics and have a rudimentary abandoned cart flow. The SEO is underdeveloped - handloom saree keywords are competitive and we don't have the content volume yet to rank for anything meaningful.

Where AI Comes In

I'm building AI into Ethlore's operations through Geenie, my personal AI OS. Current integrations:

Content Marketing: The same pipeline I use for my personal blog generates content for Ethlore - educational pieces about handloom techniques, weaver profiles, care guides, regional textile traditions. This content serves two purposes: SEO (long-tail informational keywords in the handloom space have low competition) and brand trust (customers who understand what they're buying are more likely to become repeat buyers).

Product Description Generation: We have a pipeline that takes raw product attributes (region, weave type, fiber, dimensions, motifs) and generates product descriptions for the Shopify store. The descriptions are edited before publishing, but the scaffolding saves Shuvrangshu significant time on catalog management.

What's Coming: Inventory alert automation (trigger reorder when a fast-moving SKU drops below threshold), demand forecasting (predict which products will sell in the next 60 days based on seasonal patterns and browsing behavior), and customer segmentation (identify which buyers are likely to be repeat customers and treat them differently in marketing).

The Honest Numbers

10-50 orders per month sounds like a wide range. It is. We have good months and slow months with no obvious pattern yet - we don't have enough volume to distinguish signal from noise in the sales data. The average order value is in the $80-200 range depending on the product. We're profitable on a per-unit basis. We're not profitable when you factor in the time Shuvrangshu is investing.

The ₹1000 crore vision we have for Ethlore is genuine, but it's a 10-year vision, not a 3-year one. Right now we're in the phase where the goal is to build something that works repeatably at small scale before we think about growth.

Running a D2C Brand While Working Full-Time at a Fortune 500

The honest answer to "how do you manage it" is: imperfectly. There are weeks where Ethlore gets 20 hours of my attention and weeks where it gets two hours because a major deliverable at HCLTech is consuming everything. The brand has suffered from that inconsistency.

The division of labor helps: Shuvrangshu handles anything operational that requires presence in India. I handle anything that can be done remotely - strategy, tech, marketing planning, analytics. But the operations-strategy boundary is porous in practice, and there are frequent gaps that fall in between.

What I've learned: you cannot build a D2C brand effectively part-time if the operations depend on real-time judgment. Shuvrangshu being full-time is the only reason Ethlore still exists. If both of us were working on it part-time, it would have died by month six from the compounding weight of delayed decisions.

What's Next

The next six months are focused on three things: improving the visual brand (professional photography of the product line), building the content marketing flywheel (getting to 50+ indexed pages before thinking about performance marketing), and closing the first Shopify-to-Geenie integration loop (inventory management).

The longer vision is genuinely exciting to me: a D2C brand where the weavers have visibility into what's selling, where demand signals travel back through the supply chain in real time, where a weaver in Varanasi can know that their specialty - say, a particular motif style - is trending among customers in California. That's not just a business. That's a different relationship between craft and market.

We're nowhere near that. But the path there goes through the unsexy work we're doing now. And if you're interested in the intersection of craft, supply chains, and AI - or if you're a weaver looking for a direct relationship with a buyer - I'd love to hear from you.

You might also like

Further Reading