Our Investment in Tangram Vision

We know that the future belongs to intelligent machines that take care of manual tasks, freeing up our time to be more creative, to invent, and ultimately make society better.

Perception is a foundational element in robotics, but perception algorithms and sensors require highly specialized knowledge to enable, and are generally hard to design. Today’s perception code is messy and lacks abstraction layers. Sensors that are used by robots do not interoperate well, and can often fail when deployed in real world conditions. The data that sensors capture is often noisy and needs a lot of cleansing and normalization before it can be processed by a robot’s operating system.

Brandon and Adam from Tangram Vision have set out to change that. The duo previously met while working at Occipital (creators of the Structure Sensor, a ubiquitous 3D sensing platform), where Brandon was Principal Manager of the robotics and calibration team, and Adam was GM and VP of Marketing.

With Tangram Vision, Brandon and Adam have a bold mission to simplify and democratize computer vision. Much like Twilio made telephony simple and accessible via elegant APIs, Tangram is turning every engineer into a computer vision engineer by creating an API and an abstraction layer to streamline working with perception sensors and algorithms.

‍

‍

‍

‍

Today, Tangram is specifically focused on sensor integration and uptime, a messy and costly part of building service robots, industrial robots, and drones. Their v1 product will be available in early Q1 of 2021 and will help with data ingestion, cleansing, auto calibration and normalization, ultimately helping customers deploy vision-enabled robotics faster, and more cost effectively.

In the future, Tangram will continue to wrap and abstract difficult and costly pieces of the stack, will help with data storage, analytics, and will become a platform where vision algorithms can be easily plugged in, and where engineers can focus on the application layer instead of worrying about the infrastructure.

We are delighted to back this talented team and to partner with Dynamo Ventures and Trucks VC to democratize computer vision engineering.