Deco
deco provides an AI-native platform that operates and grows e-commerce storefronts through a combination of open-source software, autonomous AI agents, and senior engineering services. The platform includes Deco Studio, an AI operating system with an agentic CMS that lets non-technical team members ship site changes directly, and AI agents that autonomously perform tasks like reordering product pages by predicted revenue. The company operates on a results-based model where customers pay for measurable growth rather than software licenses.
- Artificial Intelligence
- AI Agents
- Retail Technology
- Commerce & Shopping
- Software Only
- Marketing Technology
Funding
Founders
Founder details are not available yet.
Product
Problem
E-commerce teams face a production gap where AI tools that perform well in controlled settings fail to deliver repeatable, auditable, and cost-predictable results in mature production environments. Site maintenance and optimization requests—such as content updates, CRO tweaks, and component changes—often get bottlenecked in development queues, delaying time-to-market for weeks or months. Additionally, enterprises struggle to operationalize AI across their organizations due to fragmented tooling, inconsistent governance, and a lack of context infrastructure that prevents agents from accessing the full picture of enterprise knowledge.
Solution
deco provides an AI-native platform that operates and grows e-commerce digital experiences through a combination of open-source software, AI agents, and forward-deployed engineers. The platform includes Deco Studio, an AI operating system that serves as an open-source workspace and control plane, and an agentic CMS that allows non-technical team members to ship site changes directly without waiting for development tickets. AI agents autonomously perform tasks such as reordering product collection pages by predicted revenue, diagnosing site performance issues, and generating content, while senior engineers handle complex implementations and ongoing optimization. The company operates on a results-based model where customers pay for measurable growth rather than software licenses, with AI agents and engineers embedded in the customer's workflow to continuously improve site performance, SEO, and conversion rates. The platform supports self-hosted deployments on customer infrastructure with their own keys and data, or a managed service operated by deco.
Target Audience
Primary customers are e-commerce brands and retailers—particularly those operating on platforms like VTEX—that need to continuously optimize their storefronts, along with enterprise teams seeking to operationalize AI across their digital operations without building internal AI platforms from scratch.
Features
- Open-source core platform with self-hosted deployment option, allowing customers to run the software on their own infrastructure with their own keys and data
- Deco Studio workspace featuring AI agents that can autonomously execute tasks, diagnose site issues weekly, and run automated workflows with context, permissions, auditing, and cost tracking in one place
- Agentic CMS that enables non-technical users to edit content, create new components, and publish changes directly through natural language instructions with Git-based safety and design system enforcement
- MCP (Model Context Protocol) integration for secure, controlled access to Google Analytics and Google Search Console data via OAuth, enabling AI agents to generate reports and automations with scoped permissions
- Autonomous agents that learn from user behavior, retrain on configurable cadences, and publish changes directly to production—demonstrated to generate +10% revenue on collection pages for FARM Rio in controlled A/B tests
- Monitoring dashboard with 24/7 observability of errors, funnels, and site health, including page views, cache hit rates, and performance benchmarks
- AI agents integrated with Google Analytics and Google Search Console through MCP, supporting automated diagnostics, SEO rewrites, and catalog improvements like alt-tag generation