Loops raises $3.5m seed to build unified AI layer for retail commerce

A a16z-backed startup targets fragmented retail AI stacks with a single operating system, with financial services and energy verticals next.

By Disrupts Editorial Team Wed, 21 October 2026 AI level 2 of 5

AI level 2 of 5: drafted by our AI editorial assistant from the company's announcement; fact-checked, edited and signed off by Mark Walker, Editorial Director. What the levels mean

A brightly lit data center aisle features a suspended black server rack in the foreground, flanked by rows of other black server racks, all displaying numerous blue and green indicator lights.

Loops, a San Francisco-based enterprise AI startup, has closed a $3.5 million seed round to accelerate what it frames as an "AI operating system" for e-commerce, with ambitions to replicate the model across financial services and energy. The round was led by a16z Speedrun and includes NTT Docomo, Connecticut Innovations, Alumni Ventures and Monday Capital.

The funding announcement arrives at a moment when retailers face a crowded and fragmented AI vendor landscape. Separate tools handle search optimisation, product catalogue enrichment, virtual try-on, and customer support, with the retailer left to integrate and maintain each system. Loops positions itself as the layer that consolidates those functions around a shared understanding of product data, business context, and customer intent.

Consolidation over features

The company's architecture works by treating product knowledge as a shared substrate. Enriched catalogue data improves AI-powered search; search activity feeds context to a shopping assistant; customer conversations surface gaps in product information, which in turn improves future enrichment. Each module reinforces the others rather than operating in isolation.

Early deployments include Turkish supermarket chain A101 and fashion brand Mavi. Loops reports that sessions involving its platform at A101 showed roughly 83% higher observed purchase conversion compared with a control segment, though the company notes the figure is observational rather than a randomised controlled measure. Across its broader pipeline, the startup says 64% of pilots have converted to paying customers, across a base of more than 70 brands and pilots handling millions of shopper interactions. Those numbers are company-reported and unaudited at this stage.

"The biggest mistake we see is treating AI as a collection of features," said Ilker Selim Zorluoglu, President and Co-Founder of Loops. "A retailer does not need five different AI products that each understand one slice of the customer. It needs one system that understands the catalog, the business and what the shopper is trying to do."

The vertical AI operating-system play

The retail wedge is deliberately narrow. Loops is explicit that commerce is a proving ground for a broader architectural thesis: that enterprise AI value accrues to vertically integrated systems carrying domain knowledge, workflow integrations, and agentic controls, rather than to general-purpose foundation models with thin application wrappers. The company names financial services and energy as its next target verticals, and says it is building specialised small-language-model capabilities for deployment environments where a focused, domain-trained model outperforms a larger generalist one.

That thesis sits squarely within a wider strategic debate playing out across enterprise software. Incumbent SaaS vendors from Salesforce to SAP are racing to retrofit AI agents into existing workflow layers, while a cohort of vertical-AI startups argue the better architecture is built domain-first. The Loops model is closer to the latter camp, and the a16z Speedrun backing signals conviction that the vertical-first approach can compound into defensible platform positions.

For cross-sector capital allocators, the implications reach beyond retail. If the vertical AI operating-system model scales from commerce into financial services, it creates pressure on core banking software vendors and wealth management platforms already navigating their own agentic AI transitions. In energy, the same architecture could challenge the industrial IoT and SCADA-layer vendors that currently own operational data in grid and upstream contexts. Loops is a seed-stage company and those ambitions remain unproven, but the architectural bet it is making in retail is precisely the kind of cross-vertical convergence play that will determine which AI infrastructure layer wins in each regulated industry over the next decade.

The $3.5 million round is modest relative to the ambition stated, and the company will need to demonstrate repeatable, verifiable performance metrics as it moves into more complex and regulated verticals. The next meaningful signal will be whether the financial services pilot programme produces results comparable to the retail deployments, and at what funding level the company pursues a Series A.