What is Pipeless Agents?
Pipeless Agents can be quickly enabled in three steps: create a project (each project includes an intelligent agent that can sync with a code repository and multiple video-processing pipelines), connect a video source (compatible with IP cameras, online streaming, etc.), and customize agent logic (requiring only two lines of code to invoke Webhooks, write to databases, or send messages). The platform offers a pre-built agent catalog covering scenarios such as PPE monitoring, table-occupancy statistics, face blurring, checkout queue analysis, customer age grouping, and people recognition. All video frames are analyzed instantly and then immediately discarded, with support for custom region processing and end-to-end encryption to ensure privacy compliance.
Key features
Supports industrial-grade PPE monitoring: it sends real-time notifications when workers fail to wear designated personal protective equipment, with customizable PPE types;
Provides table occupancy monitoring: accurately tracks the duration customers spend at restaurant tables and supports exporting data to Google Sheets or any data aggregation platform.
Use cases and audience
Targeted at enterprise users who require automation tasks based on visual input, including developers and IT decision-makers in fields such as industrial safety, retail operations, smart offices, smart cities, and environmental monitoring.
- Industrial scenario: detects whether workers are wearing personal protective equipment (PPE), with support for customizing the required PPE types.
- Catering scenario: Monitoring the duration customers spend at their tables in a restaurant
- Retail scenario: Exporting checkout queue duration data to a data aggregation platform or Google Sheets
Video stream input, supporting various sources such as IP cameras and online streaming platforms.
Structured data and event streams, which can be forwarded to target systems like Webhooks, messaging systems, MongoDB, and more.
Frequently asked questions
What is an Agent?
The Agent is a processing loop that receives structured data from the video pipeline and can perform arbitrary programming operations (such as calling Webhooks, writing to databases, etc.).
Is my video privacy protected?
Yes. Video frames are analyzed immediately upon arrival and then discarded right away, with no recording taking place; it supports end-to-end encrypted transmission and can be deployed on-premises within the customer’s own infrastructure as needed.
How to use
- 01
Creating a project: Each project includes one Agent and several video pipelines, with the Agent synchronously processing all pipelines
- 02
Connecting video sources: Supports various input sources such as IP cameras and online streaming platforms
- 03
Configuring filters: The system automatically extracts structured data and events from the video stream, which the Agent then receives
- 04
Updating Agent logic: With just two lines of code, the Agent can start receiving data, supporting operations like invoking Webhooks, sending messages, and storing data in databases
Official website & sources
Sources
Updated on 2026-10-30
