Loops Raises $3.5M Seed to Connect AI Shopping Workflows

Loops is asking retailers to let a conversation carry a sale across several departments. A customer's question can touch search, product information and service before anyone decides what to buy, leaving the retailer responsible for keeping the answer useful as it travels.

The San Francisco commerce AI company announced a $3.5M Seed round on October 12, 2026, backed by a16z speedrun, NTT Docomo, Connecticut Innovations, Alumni Ventures and Monday Capital. Product development, customer deployment capacity and international expansion are the stated uses of the funding. That middle allocation matters: a shopping assistant has to learn the business behind the catalog before it can help sell what is inside it.

Loops combines AI search, shopping assistance, marketplace support, product enrichment and virtual try-on. Its proposition gives retailers a shared foundation for these activities, making the quality of their product knowledge part of the customer experience. The financing supports an attempt to turn that relationship into software that works across more retailers and markets.

The purchase that supplied the brief

The founders' account starts with an abandoned gift purchase. İlker Selim Zorluoğlu wanted to buy something for his daughter, asked the retailer for information and waited for help that arrived too slowly. The purchase fell away. Ari Nazir and İlker Selim Zorluoğlu began working on the larger problem that the episode exposed: a person can express a need clearly while the store's systems make answering it unnecessarily complicated.

The a16z speedrun company directory identifies Ari Nazir as CEO, İlker Selim Zorluoğlu as President, Yusuf Bahadır as CTO and Hakan Bas as a cofounder. It lists Loops as founded in 2025 and part of cohort 006. The company brings that team into a commercial problem with an unusually familiar customer behavior: people ask questions while deciding whether to spend.

A retailer can have the right product without giving the customer enough confidence to choose it. Search, catalog administration and customer service each have useful information, but the buyer experiences the time and effort needed to connect them. Loops is building around the decision inside that interval, when interest is still present and the sale remains optional.

The catalog becomes part of the conversation

Loops' architecture starts below the shopping assistant. Product enrichment makes descriptions and attributes easier for software and customers to interpret. Search can then use a shopper's stated intent, while the assistant continues the conversation with access to the same product context. Virtual try-on adds another way to evaluate an item before buying.

The company's product site presents these tools as connected parts of its commerce system. It also lists integrations with Shopify, WooCommerce, Akinon, ikas, Ticimax, IdeaSoft and T-Soft, alongside its script and APIs. These are vendor-described capabilities; buyers still need to establish how a deployment will work with their own catalog and operating processes.

That makes product information a sales input with a more visible destination. A missing attribute can surface in the assistant's answer, and a useful customer question can reveal what the catalog never explained. The commercial implication is a closer relationship between the people maintaining product data and the people responsible for the buying experience. A fluent conversation places more attention on the information carrying it.

Retail deployments give the financing a destination

A101 and Mavi are named customer examples in Loops' announcement. A101 uses the platform for catalog navigation and shopping assistance, while Mavi's deployment covers AI search and help with preferences such as style, occasion and fit. Those settings put the product in front of different purchasing decisions rather than limiting it to a single demonstration catalog.

Loops reports more than 70 customers and pilots combined, with 64% of pilots converting to paying customers. The figures are company reported, and the combined count includes trial deployments. A retailer moving from a pilot into a paid relationship is nevertheless a specific commercial milestone: someone has chosen to keep the system inside the buying experience after trying it.

Customer outcome claims require closer reading. Loops' announcement describes an A101 purchase-conversion comparison as observational, so readers should avoid treating the comparison as proof that the software alone caused the result. The broader deployment evidence supports a company working with named retailers; it leaves the measurement of incremental value as a separate task for each buyer.

Deployment work is inside the capital plan

Funding customer deployment capacity puts implementation inside the investment case. A retailer needs its products, business rules and service information to remain intelligible when a shopper moves between activities. Expanding the software therefore brings work around the catalog, the integration and the people accountable for the answers, alongside work on the models themselves.

International expansion adds another dimension to that responsibility. Retail language includes local expectations about sizing, product descriptions and service. Loops' commerce system will encounter these differences as it enters new markets, making the deployment experience part of how the company learns where the product needs to change. That is an operating implication of the expansion plan, rather than a claim that every international requirement is already solved.

Loops also describes financial services, energy and specialized small language models as future directions. Those ambitions extend the idea of software built around an industry's knowledge and workflows. Commerce supplies the present customer examples, while the broader ambitions remain plans; the new financing does not establish commercial outcomes in those other sectors.

A retailer still owns the answer

For a retailer evaluating Loops, the choice reaches beyond whether a chatbot can produce a pleasant sentence. The buyer is deciding how search, assistance and product information will share context, and who will correct that context when a customer asks something the catalog does not explain. Bringing the capabilities together changes where that responsibility can be exercised.

The $3.5M round gives the founders resources to develop the product and support more of those deployments. It leaves the company close to the original question that started the work: how quickly can a store become useful to someone who already wants to buy? As Loops enters another retailer's catalog, the answer will depend on the product information, service knowledge and implementation work that travel with it.