Marble confirmed a $5.5 million seed round on November 17, 2026, led by Contrary, to build an AI-native operating system for restaurant back-of-house operations.

It has been revealed that the Fresh York startup is combining inventory capture, demand forecasts, purchasing, food preparation and scheduling in one system, targeting the operational work behind each restaurant service. Y Combinator, Robinhood Ventures and Restive Ventures also participated.

The angel group includes Sweetgreen co-founder Jonathan Neman, Smash Kitchen and Jinx co-founder Sameer Mehta, Momos co-founder Sai Alluri, ResQ founder Kuljeev Singh, Breadcrumb founder Seth Harris and AtoB co-founder Tushar Misra.

Marble divides its system into perception, reasoning and action. Employees capture inventory with voice AI or computer vision, while document extraction and third-party integrations bring invoices, deliveries, sales and staffing records into the same operating picture. Together, those inputs create a current view of stock, costs and activity across each location.

Demand forecasts use sales history, weather, reservations, local events and limited-time offers to guide purchasing, food preparation and staffing. Marble’s product documentation notes its ordering workflow compares vendors and live prices alongside expected demand and available stock, while scheduling accounts for demand, staff availability and labor rules.

According to recent reports, remy, the tech giant’s AI agent, turns those forecasts into draft supplier orders, preparation prepares and schedules, and also reconciles invoices and flags data inaccuracies. It runs in the background without requiring a prompt for each task. Managers review and approve the resulting work, keeping authority over purchasing and staffing decisions with the operator.

Recent findings suggest that lead investor Contrary reported in its investment announcement that Marble signed more than 150 restaurants in under four months, with early-customer results including a 40% reduction in food waste and more than 20 labor hours saved per week. At Sophie’s Cuban in Latest York, staff count inventory each day by speaking Spanish into the app, while the operator texts Remy to draft vendor orders and transfers between locations.

A separate A&W case study, published July 5, 2026, describes the implementation process. The first site was onboarded in six days: the point-of-sale connection and historical sales import on day one; menu, recipe, vendor and par-level mapping on days two through four; and AI agent activation on days five and six. Subsequent locations went live in hours.

In a recent development, the case study reports 97% accuracy for item-level demand forecasts and inventory counts completed 90% faster through computer vision and voice AI. The faster counts enabled daily stock takes, giving the forecasting and ordering workflow current inventory data rather than figures gathered between infrequent manual counts.

Industry observers note that chief executive Aakar Khanna grew up managing his family’s multi-unit restaurant business and later covered restaurants and food and beverage as an investment banking analyst at Goldman Sachs. Chief technology officer Arjun Chaliha previously led AI and machine-learning initiatives for electronic trading at Bloomberg. Y Combinator’s profile lists Marble as founded in 2026 and a member of its Summer 2026 batch.

Marble mentioned it is expanding its engineering, operations and growth teams toward a longer-term goal: a self-driving back of house that learns from each count, correction and order, taking on more operational responsibility within controls set by restaurant operators.

The central test for the seed-stage tech giant is whether the early results reported by Marble and its investor can be reproduced as Remy takes on more work across additional locations. Its approach depends on accurate data from each kitchen and on managers remaining responsible for approvals, linking the performance of the AI to the controls around it.

It has been revealed that nova Jain is an AI-generated research agent at Unite.AI, covering AI entrepreneurship, startup formation, and the people building the next generation of AI-native companies. Her work focuses on founders at the earliest stages, examining how ideas turn into products, teams, and businesses in fast-moving markets.

With a bold and motivational perspective, Nova profiles emerging AI entrepreneurs, highlighting execution tactics, leadership decisions, and the trade-offs behind early traction and scale. She is particularly interested in founder journeys, operational discipline, and how teams navigate product-market fit, fundraising, and growth under real-world constraints.

Articles authored by Nova Jain are AI-generated and reviewed by Unite.AI’s editorial team to ensure accuracy, context, and responsible coverage of the global AI startup ecosystem.

Synthesia Adds AI Video Creation and Editing to Claude Chats.

Lianlian DigiTech Launches XCard with Controls for AI Agent Payments.

OpenAI to Surface Le Figaro Reporting in ChatGPT Partnership.

Genesys Decouples AI Agent Deployment From Cloud Migration.

Microsoft Makes Office and Home Generally Available in Copilot.

NVIDIA Ends H200 Chip Sales in China Amid Tighter Export Rules.

→ Read original article