GeoComputing and Leostream Deliver High-Performance Geoscience Computing From Anywhere
Partnership centers on RiVA platform, secure remote access, and a shift toward private cloud
In a recent Leostream podcast, GeoComputing executives Jay Kirby (VP of Strategic Solutions) and Eric Johnson (Director of Consulting Services) joined Leostream CEO Karen Gondoly to discuss their long-running partnership and how it's reshaping computing for the oil and gas industry.
GeoComputing's RiVA platform, now in its seventh generation, gives geoscientists secure access to graphics-intensive applications like Petrel, used to analyze seismic data and 3D geological models. Leostream provides the secure access layer that lets users connect from anywhere, while technologies like Mechdyne TGX preserve an in-office experience for remote users when latency stays under 100 milliseconds.
Because systems are pre-built and integrated before shipping, GeoComputing has cut deployment times by about 70% compared to the traditional 18–36 month timeline. The platform has also enabled global collaboration — one Houston-based customer with teams in Denver, Houston, and Calgary centralized data across all three offices, part of the hundreds of petabytes GeoComputing has migrated across its customer base.
Security is built in at every layer, with zero-trust network segmentation, enterprise identity integration (Active Directory, Entra ID), and Leostream-managed authentication and single sign-on — ensuring sensitive data never touches a user's endpoint.
On AI, Kirby said the industry is still early. GeoComputing is focused on internal use cases, like smarter customer support, and helping customers organize existing data, while preparing infrastructure with the latest NVIDIA GPUs for AI-enabled applications. Leostream allows customers to create dedicated AI resource pools rather than upgrading every workstation, controlling access and cost.
The company has also shifted from public cloud back toward a private cloud model, after finding public cloud deployments often cost two to five times more than expected with inconsistent performance. Private cloud is now central to GeoComputing's strategy alongside traditional on-premises deployments.
Looking ahead, GeoComputing is building faster infrastructure to support shrinking decision windows and rolling out new features like on-demand data rooms for joint ventures. "Our roadmap is driven by listening to customers," Kirby said.
More information is available at geocomputing.net. Listeners can hear the full conversation on the Leostream podcast, "The Future of Geoscience Computing: Secure Access, AI, and Private Cloud."
