Intuitive Machines deploys mission-critical software 4x faster with GitLab

Intuitive Machines is managing four concurrent lunar lander programs and an orbital transfer vehicle, alongside five satellites for their Lunar Data Network and an expanding ground network footprint supporting near-space communications infrastructure around the world.

With growth comes new complexity: more repositories, submodules across shared codebases, stricter regulatory requirements tied to national security programs, and an AI-curious engineering team that needed a safe, controlled way to adopt AI tools in an International Traffic in Arms Regulations (ITAR)-compliant environment.

"We went from running a single mission at a time to doing multiple missions, multiple teams, multiple programs in parallel," says Brian Butcher, Software Group Lead at Intuitive Machines. "Now we have satellites that we’ll have to maintain and push software updates to continuously throughout their lifespan, and that’s a completely different operational model."

For a team running four concurrent missions and two programs across dozens of repositories, governance isn’t just a convenience. Policies are what make consistent, compliant development at scale actually possible.

“I’m able to set one policy at the top, and it cascades down to multiple projects,” says Dami Alugo, DevOps Lead at Intuitive Machines. “Instead of going in and doing each individual project and remembering the configuration of each one, we just use GitLab policies to automatically enforce the rules across the board.”

Intuitive Machines’ CI pipeline usage has tripled in the past two years, and it’s not just about smoke tests anymore. Teams across the organization are using GitLab pipelines to deploy to Mission Control, to vehicles in Flatsat labs, and to kick off automated nightly testing runs. Deployments that used to take up to four hours now take 45 minutes. The team has shifted its focus from executing deployments to verifying them.

“All deployments are automated with GitLab — one push of a button, 45 minutes total. We’re now focused on the verification of the deployment, not the deployment itself. That’s a huge shift that GitLab made possible," says Alugo.

Merge request volume has increased 65% year-over-year, with 52% more merge requests now subject to formal approval rules — a signal of both faster development velocity and a maturing code review culture. The team has also adopted submodules across shared codebases, enabling code reuse across programs in ways that weren’t possible when each mission was fully siloed.

GitLab Duo is a daily tool for a growing portion of the Intuitive Machines engineering team, and GitLab Duo Agent Platform is where the team’s AI ambitions are headed next.

The journey has not been a straight line. When AI tools began proliferating across the industry, Intuitive Machines faced a challenge common to aerospace and defense companies: How do we give developers the productivity benefits of AI without exposing ITAR-sensitive code or intellectual property? The answer: Intuitive Machines runs GitLab Duo Agent Platform in a self-hosted, bring-your-own-model configuration using AWS Bedrock.

“There can be serious consequences if you put the wrong information out there,” says Butcher. “The only way to protect our intellectual property and ensure all our data is 100% secure is to contain it internally, on-prem. Without this option from GitLab, we wouldn’t be doing AI at all.”

New engineers are now onboarded directly onto GitLab Duo rather than other AI tools, eliminating the security risk of unsanctioned AI adoption while giving them immediate access to AI-assisted development from day one. “We don’t have to restrict what developers do with our on-prem resources because we know the data is secure,” says Alugo. “GitLab Duo also makes onboarding faster. New developers can review code and understand what’s going on in just a day or two.”

The GitLab team visited Intuitive Machines’ Houston office to run a hands-on GitLab Duo Agent Platform workshop, directly accelerating adoption across the engineering team. “Working with GitLab feels like a true partnership where our feedback actually helps shape their roadmap,” says Butcher. “GitLab is invested in understanding our use case and wants to help us succeed.”

Adoption of GitLab Duo has more than doubled year-over-year, growing from 80 active seats to more than 200. The AI use cases that have taken hold are practical and immediate. Developers are using GitLab Duo during merge requests to catch bugs early, reducing multi-round review cycles. They’re also using it for merge request documentation and unit test generation in VS Code, writing boilerplate that used to consume hours. One developer built an entire in-house scheduling GUI (a tool that maps into Jira) using GitLab Duo Agent Platform to generate the interface code. Alugo’s team is also using GitLab Duo to generate automated weekly and monthly developer performance summaries that are pushed directly to Confluence, giving leadership visibility into team activity without manual reporting overhead.

Butcher sees the biggest immediate AI opportunity in test and verification, an area where aerospace teams spend enormous time running tests manually. "If we can get to a point where AI helps us with regression testing, that’s a really compelling selling point to leadership," he says. It’s also where the team’s forward-looking Gitlab Duo Agent Platform ambitions are focused.

As Intuitive Machines continues to scale, security is becoming a bigger focus, driven by the company’s expansion into new mission types with fundamentally different security profiles.

Early lander missions were largely air-gapped. Software that lived on a spacecraft didn’t face the same threat landscape as software connected to public networks. But that’s changing fast. “In the early days, the company was focused on building a lander. Now we’re doing landers and communication satellites, ground networks, near-space network services — a distributed network around the world to open up communications for lunar, NASA, and commercial customers. This requires a different level of security entirely,” says Butcher.

SAST is now enforced across all pipelines via GitLab group policies that are set at the top level and applied automatically to every project before development begins. “SAST is highly relevant to the work we’re doing for the Lunar Data Network, and GitLab’s security capabilities will be essential for what’s coming next,” says Butcher.

The vulnerability dashboard gives teams visibility into security findings across the entire codebase in a way that simply didn’t exist before. They’ve also implemented an alerting framework that routes GitLab signals into Microsoft Teams, enabling the DevOps team to identify and address issues before they surface elsewhere in the organization. The team is also evaluating GitLab Secrets as a potential replacement for their existing OpenVAO integration.

According to Butcher, GitLab’s compliance and audit capabilities are increasingly important as the organization pursues Department of Defense work. "There are a lot of capabilities within GitLab that are going to help automate reporting structures and logs to prove for audits that we’re meeting compliance," he says. "That’s valuable, and it’s going to get more valuable."

Intuitive Machines’ strategic direction of Build, Connect, Operate, demonstrates that not only can it build spacecraft, but is actively building the infrastructure to connect and operate them.

“Using tools with proven processes around software development matters in this industry and to customers, and GitLab provides that level of trust,” says Butcher.

IM-3, a lander mission carrying NASA and commercial science payloads, is slated for Q4 2026, with IM-4 following approximately a year later. GitLab will support CI/CD pipelines, branching strategies for mission freezes, configuration file reviews across subsystem teams, and real-time mission response capabilities, just as it did for Odysseus.

The Lunar Data Network, a persistent communications infrastructure for cislunar space, anchored by five Altus relay satellites currently in development, is Intuitive Machines’ primary forward-looking focus. It’s also the project where they expect to use GitLab’s full platform stack most completely: Ultimate security features, Duo Agent Platform, continuous deployment, and SAST across an interconnected ground and space network. The team is also interested in exploring GitLab Orbit. As code reuse expands across programs and submodule complexity grows, the ability to navigate relationships between requirements, source code, and metadata across the entire repository ecosystem has clear appeal.

Intuitive Machines was the first commercial organization to land on the Moon. The next chapter — building the infrastructure for humanity’s permanent presence in space — will be written the same way: with GitLab as the platform powering every line of code.