Geopipe – Building realistic 3D models of the world with machine learning

Geopipe – Building realistic 3D models of the world with machine learning


Name: Geopipe

Legal Name: Geopipe, Inc

Location: New York City, New York

Founded: 2016

Founder(s): Christopher Mitchell and Thomas Dickerson


Social Media Following: 104 Likes on Facebook, 199 Followers on Twitter, 108 Followers on LinkedIn 

INDUSTRY – 3D Computer-aided Design Software

Size: $8.99B in 2018 (Source: Grand View Research)

Projections: $13.04B by 2025, at a CAGR of 6.2% (Source: Grand View Research)


Geopipe develops algorithmic tools to create 3D models for realistic and expansive virtual worlds for gaming, architecture, real estate, urban planning, and other use cases. 

The Product 

Geopipe develops algorithms to create highly detailed 3D virtual models from a large amount of 2D and 3D data. Its proprietary system analyzes data, including satellite images, maps, laser scans, etc. and then applies machine learning to instantly identify correlations between data sources, understand the structure of data, and create comprehensive 3D models at varying scales. The company currently offers 3D models for cities such as San Francisco, New York, Boston, Sydney, Seattle/Bellevue, and Salt Lake City

Geopipe differentiates itself from other companies within the same domain by offering more semantic modeling capabilities through its system rather than providing similar solutions as their counterparts. The semantic modelling feature also offers a large amount of content and instantly correlates models to real-world data with minimum manual effort. Geopipe makes its models accessible to customers who can easily customize them using their own software suites without affecting the high-level visual quality of the original models. 

Geopipe’s models are compatible with popular software used by professionals, from 3D design software to gaming or virtual reality engines like Unity and Unreal. Users can also export mesh models in multiple formats such as FBX, OBJ, DAE, OSG, and GLTF, among others. 

Origin and Founding Team 

Geopipe was founded by Christopher Mitchell and Thomas Dickerson in 2016. 

Christopher Mitchell is the CEO of Geopipe and holds a Ph.D. in Computer Science from New York University. In 2000, Christopher founded Cemetech, an online community for learning and sharing knowledge on programming and developing hardware. 

Thomas Dickerson is the Chief Science Officer of Geopipe and holds a Ph.D. in Computer Science from Brown University. His past professional experience predominantly includes working as a researcher or research assistant on various programs at a number of institutes, including Princeton University. 

Christopher Mitchell and Thomas Dickerson have known each other since 2002. Passionate about gaming and programming, they were both independently exploring projects to recreate the real world for video games between 2010 and 2012. The idea of replicating the real world in a virtual space fascinated both of them, and they decided to work together on developing a solution that could do that using the latest developments in the programming space. In 2016, they co-founded Geopipe to create immersive, expansive virtual models of the real world, automatically and with minimum human intervention. 

Performance and Trends

In 2017, Geopipe was selected for the Techstars NYC accelerator. In May 2019, the company received a National Science Foundation (NSF) Small Business Innovation Research (SBIR) Phase II grant worth $750k to help them with commercializing its technology.

Subscribe to our newsletter

Load more