From traditional craftsmanship to technology: how an Uzbek carpet weaver launched an artificial intelligence startup

Hikmatillo Umarhojiyev, San Francisco, CEO, LinkedIn

How It All Began

I am a tech entrepreneur and a hand-knotted carpet artisan. I started my career in traditional industries—carpet weaving and textile trade. For over 15 years, I worked in craftsmanship, retail, and customer relations, growing a business in a sector that had changed very little over many years.

Over time, I realized that the industry’s core problems were not artistic or product-related, but structural. Valuable carpets were sold using outdated models; customers felt insecure before visiting stores, and consultants spent much of their time on repetitive and inefficient interactions. This slowed growth, wasted human potential, and hindered the expansion of traditional businesses.

I consciously decided to transition into technology without leaving the industry. My goal was to understand how artificial intelligence could be responsibly applied to solve the long-standing problems of traditional business. I learned to use modern AI systems not to replace craftsmanship, but to turn advanced technologies into practical tools for retailers.

This path—transitioning from traditional entrepreneurship to technology and bringing technology back to traditional industries—became the foundation of my work. Combining deep industry experience with AI, my co-founder Shavkat Homidov, who is responsible for technical architecture and systems engineering, and I created the Findecor platform. It helps understand the client’s space, their desires, and intentions before they visit the store, transforming the process of selecting and purchasing carpets while preserving the value of the human experience.

Launching the Startup

The company emerged not on a theoretical basis, but through direct testing in the field. I developed the initial AI prototype and tested it with real carpet sellers using live inventory data. In the first stage, I personally handled the product concept, customer acquisition, and pilot demonstrations, while Shavkat Homidov was responsible for system architecture, AI infrastructure, and technical implementation.

We consciously avoided rushing into rapid expansion; instead, we grew step by step, keeping pace with traditional stores in the US. The team was also formed gradually: first, AI engineers and architects joined, and as the platform improved, product and business specialists were added. This approach allowed us to address fundamental industry problems before expanding functionality.

Target Audience

The startup is designed for carpet and carpet product sellers, particularly medium and large stores with a wide assortment and a sales team.

The platform provides the following capabilities:

— analyzing customers’ rooms and preferences before store visits
— recommending relevant products based on real inventory and sales priorities
— reducing repetitive consultations
— increasing conversion while preserving the value of human expertise.

Today, the platform has been tested in several stores in the San Francisco area—Menlo Park, Palo Alto, and San Francisco. Currently, we are negotiating with about 30 new partners and plan to reach 300 active customers by the end of 2026.

Incubation Programs

The startup was selected and supported by several international innovation programs. It is part of the Silkroad Innovation Hub and the AlchemistX Silicon Valley Residency, which are focused on global technology companies.

Additionally, we collaborate with Astana Hub and IT Park Uzbekistan, reflecting our position at the intersection of technological innovation and traditional industries in various regions. These programs played a crucial role in confirming the product’s relevance in the US market and presenting it as a cross-border solution.

Challenges

One of the main challenges was integrating AI into a deeply traditional industry. Many sellers initially perceived AI as a threat to the human experience. In pilot tests, we demonstrated that AI works best as a support tool that assists consultants rather than replacing them.

Another issue was the lack of systematic digital data in the carpet industry. Product information was often fragmented and unclear. To solve this, we developed data normalization, comparison, and recommendation systems, which became the key technical differentiator of our platform.

The main lesson: innovation in traditional sectors requires not just technology, but also trust, patience, and deep knowledge.

Achievements

A specialized AI platform has been created to address the long-standing structural problems of carpet retail. The project has been successfully tested in several US stores and selected for prestigious international acceleration programs. During this time, a growing group of partner retailers has formed in the American market, and the product has practically demonstrated how AI can modernize industries with rich heritage while preserving the value of human experience and expertise.

Plans

By the end of the year, we will focus on expanding our operations in the US, converting pilot projects into long-term partnerships, and improving the accuracy of room analysis and recommendations through AI.

In the long term, our goal is to expand the platform across the entire US and enter new global markets where traditional retail faces similar structural challenges. Our task is to demonstrate that AI can responsibly modernize heritage industries without erasing the craftsmanship, knowledge, and human considerations that define them.