Emilia AI, a restaurant intelligence startup founded in 2024 by Colombian entrepreneurs Andrés Cajiao and Sebastian Rojo, has secured a US$1.5 million pre-seed funding round led by Prosus Ventures with participation from technology-focused angel investors. Prosus Ventures contributed US$1 million while the remaining US$500,000 came from a group of angel backers. The company combines existing security camera feeds with point-of-sale data to help restaurant operators increase sales, prevent fraud, and manage staff more effectively.
How the technology works
Emilia AI connects to cameras that restaurants already have and reads visual information such as table status, customer behavior, and staff movement. That footage is combined with point-of-sale data so software agents can analyze operations in real time. The result is an operating system that supports labor optimization, fraud prevention, sales coaching, and broader control over restaurant activity.
Addressing workforce instability
Emilia AI was designed for an industry with a chronic workforce problem. In Brazil, a typical bar or restaurant replaces its entire staff roughly every 16 months, a turnover rate above 70 percent per year and more than double the rest of the services sector. The platform's coaching capability helps employees increase productivity and compensation, which the company expects will reduce staff turnover.
Two AI agents
The company has developed two primary artificial intelligence agents. Coach acts as a virtual guide that helps waitstaff offer drinks and desserts when plates are emptying, improving sales performance. Sentinel turns the same camera footage into a fraud prevention interface by flagging food leaving the kitchen without being recorded or cash not reaching the register.
Early results and client pilots
Early deployments suggest the approach can generate meaningful returns. At Alife Nino, a Brazilian restaurant group with more than 100 locations, a pilot produced roughly a 28 times return on investment, representing about US$300 in extra revenue per table each month against a cost of about US$10 per table. A separate pilot with iFood franchise client Jin Jin used Sentinel to identify unauthorized stockroom visits and checkout discrepancies, and the solution is now being scaled to more than 77 locations.
Strategic value for Prosus
Prosus Ventures views the investment as a strategic bet on deeper artificial intelligence integration across its portfolio, especially in foodservice where the Dutch group already has a strong presence through iFood. Matheus Campos, Head of Latin America at Prosus Ventures, said Emilia AI provides a unique value add for restaurant merchants and enhances point-of-sale capabilities. The move supports the group's ambition to target Brazil's restaurant and foodservice sector, projected to be worth more than US$75 billion.
Expansion and independence
The fresh capital will enable Emilia AI to continue developing its platform and expand across Brazil, where many restaurants still rely on basic point-of-sale software, and into the wider Latin American market. Although iFood and Emilia AI may test synergies, Prosus Ventures said the goal is to keep the startup independent so it can serve the entire sector. The founders selected Brazil for the launch because of the country's large restaurant labor challenges and openness to testing new technology.
Privacy and data handling
Emilia AI says the technology does not use biometrics and does not record conversations outside customer service interactions. Video footage is not stored outside the restaurant, according to the founders. The system is designed to analyze physical world information such as camera records, checkout data, and movement in the dining area that would otherwise be lost.
Emilia AI's pre-seed round provides the capital needed to scale a platform that turns everyday restaurant operations into real-time, actionable insight. By focusing on staff augmentation rather than replacement, the company aims to improve both commercial results and working conditions in a sector defined by high turnover. With early pilot returns and fraud detection gains already visible, the startup is positioning itself as a practical artificial intelligence layer for restaurants across Brazil and Latin America.