Company information · October 2026

Sangati. For the moments a restaurant floor misses.

Public company facts for editors, analysts and hospitality-industry researchers. Sangati is an early-stage restaurant-floor AI product under field validation in Mumbai.

Company overview

Sangati builds an AI-supported restaurant-floor observation layer intended to help teams track table occupancy and staff table checks. The pilot assesses compatibility with existing CCTV and local computing hardware. The goal is better operational visibility; outcome improvements have not yet been independently established.

Product / brand
Sangati™
Location
Mumbai, India
Founder
Prasad Mehta, restaurant operator and hospitality entrepreneur
Category
Restaurant operations software / on-premise computer vision
Stage
Restaurant pilot under field validation; not an independently verified commercial rollout
Website
sangati.tech
Media contact
hello@sangati.tech

What is being validated

  • Person and table occupancy observation from assessed camera views
  • Observed staff table checks and elapsed time since a check
  • Separately staff-recorded table Ready status
  • Camera/recorder compatibility, floor coverage and local device suitability

Not independently validated: detection accuracy or false-alert rates at commercial scale, bill intent, food delivery, guest satisfaction improvement, revenue impact, or kitchen bottleneck resolution. A staff visit is not proof of a completed service interaction.

Privacy and deployment distinction

The stated pilot design does not use facial recognition, personal identity inference or guest profiling. Camera inference runs locally on suitable venue hardware. Authorized remote access to operational dashboards is distinct from local camera processing; data access and retention need to be agreed per venue. See the privacy page.

Recommended editorial angles

  • Why restaurant computer vision is harder than detecting people in a room
  • What operational signals can and cannot be established from CCTV
  • The practical constraints of running AI locally on hospitality infrastructure
  • Why rigorous false-alert measurement matters before restaurant-floor automation

For current technical context, read the restaurant occupancy AI field guide and product workflow.

Media enquiries

Editors can request a founder interview, product walkthrough or permission-cleared pilot material. Publication of pilot results should be contingent on verified measurements and any required venue consent.

Contact Sangati for an interview →

Last updated: 18 October 2026. This is a first-party information page, not independent third-party validation.