How detaso Ended Environment Drift with Flox
detaso builds healthcare operations software that therapists use to document care, submit claims, and collect reimbursement. Ultimately, the company’s software is the difference between people getting the care they need when they need it and … falling through the cracks.
With so much at stake, detaso’s platform needs to behave reliably across retries, failures, and recovery events. Reliability of this kind depends on workflows that produce the same result every time they run, regardless of the circumstances. detaso’s search for a fast, secure, repeatable software delivery motion led it to Flox, a vendor that specializes in making software build and run reproducibly and deterministically, anytime, anywhere.
Drift Was Never an Option
Many companies accept drift across their software development lifecycle (SDLC) as a cost of shipping.
Given the stakes, detaso could not do this. The company was inspired by a bold, seemingly impossible vision: software infrastructure that reliably runs and behaves the same from local dev to production, on-premises or in the cloud, yielding the same results every time.
With a reliable, reproducible software backbone like this, people would get the care they need, providers would get paid, and detaso’s business could grow and adapt more easily. This same software infrastructure would enable detaso to respond faster to changing regulations; absorb industry consolidation; and more quickly launch new services, expand operations, or undertake other business activities … without the risk of disruption.
“What we wanted was a predictable, reproducible environment locally, so we knew what we were handing off,” explains detaso Principal Software Engineer Ryan Schlesinger, who introduced Flox and championed its adoption. “We hoped to eliminate the differences in each engineer's local setup, so software would then move through CI and into production without surprises.”
Container Guarantees, Without the Containers
Teams use containers to isolate and run software reliably in production. Schlesinger wanted container-like runtime guarantees in local development without the hassle and overhead containers add to everyday engineering work. He found a potential solution in Nix, open source software for reproducibly building and running software. But Nix has a well-known learning curve, and Schlesinger worried that Nix by itself would be a bridge too far for his team. Digging deeper, he discovered that most tools designed to simplify Nix still expose too much of its internals to engineers who just wanted to work with their preferred tools and workflows.
That’s when Schlesinger found Flox. Powered by Nix, Flox enables teams to build and deliver software that runs reliably and reproducibly at runtime, anytime and anywhere, without requiring any knowledge of Nix. “It seemed like magic. After I used it for awhile, I thought ‘This scratches the itch in the right way.’ I can ship environments and my developers pull and run them… and they just work. And they don't have to know how they just work,” he said.
“But also some of them … use Flox for personal projects. That's not something we ever accomplished with the other Nix ecosystem tools we tried. That was one of my criteria. It's not super useful to build yet another black box.”
It seemed like magic. After I used it for awhile, I thought ‘This scratches the itch in the right way.’ I can ship environments and my developers pull and run them… and they just work. And they don't have to know how they just work.
Ryan Schlesinger, Principal Software Engineer, detaso
No Two Machines Alike
Schlesinger is detaso's most senior engineer and its DevOps lead. When he joined, every developer was running Homebrew on macOS. No two machines matched. OpenSSL versions varied. Some engineers had migrated from x86-64 to Apple Silicon MacBooks … but were still running everything through Apple’s Rosetta emulation layer. Debugging a single new dependency could stall the team for hours.
The operational costs showed up everywhere:
- Adding or changing software dependencies meant manually debugging conflicts across every developer's machine. Changes that should have taken minutes led to hours of troubleshooting.
- Onboarding was slow because new engineers, some of them external contractors, had to manually cobble together working environments from confusing instructions. There was no single source of truth.
- Local dev, CI, and prod ran different stacks. Code that worked locally could fail in CI; code validated in CI could behave differently in production. Diagnosing failures was time-consuming and tedious.
- A single person, Schlesinger himself, was the only one who understood how detaso’s dev environments worked. If he was unavailable, the team sometimes relied on online fixes; however, these could be unreliable, outdated, or flat-out incorrect.
These weren’t just engineering “inconveniences.” Every hour lost to debugging “works on my machine” failures was time not spent building (or enhancing) payer integrations, improving validation logic for new billing scenarios, generating compliance-ready reports, or adapting to Medicaid rule changes that vary by state. Development tooling had become a huge bottleneck for detaso’s business aspirations.
From “My Machine” to “Everybody’s Environment”
With Flox, detaso's daily grind of dependency debugging disappeared. So, for example, improving a business service by adding a Python library (WeasyPrint) to an existing Rails application turned out to be fairly straightforward. Schlesinger created a declarative Flox environment that includes WeasyPrint and its native dependencies and pushed it to FloxHub, Flox's central hub for versioning environments. He told his team to run flox pull to get exactly the same environment, with exactly the same software, on their own machines. Just like that, detaso’s clients could start pulling compliance documents and financial reports faster and more reliably, instead of waiting on hidden Chrome instances to render each report.
“The big win was getting rid of the invisible dependencies that used to follow code around,” Schlesinger says. “It went from ‘I have a working machine’ to ‘Everybody has exactly the same working environment.’”
Because Flox makes the environment itself the source of truth of what software needs to run, detaso now has the ability to do something that once seemed impossible: to see, in one place, exactly what its software needs to run. So instead of guessing or scanning or inferring which software is used in an environment, detaso just reads them from each Flox environment’s declarative manifest.
“The point isn’t just faster development. Yes, we go faster now, but there also isn’t this stutter-step anymore. We’re just better able to handle different kinds of change,” Schlesinger asserts.
A Paved Path to Production
detaso didn't need Flox to replace its production stack. Schlesinger valued and understood the company's container-based deployment on AWS; he had no reason to discard it. What attracted him to Nix, and what sold him on Flox, was that it paves the path from development to production. Engineers are now able to pull and use declarative Flox environments across their systems and projects. Code now moves through local development with fewer surprises because its runtime didn’t change every time a machine updates and/or reboots. Best of all, Flox dropped right into detaso’s existing workflows with almost no friction.
Best of all, Flox gives detaso a software delivery foundation that scales. As the company grows, whether that means expanding into new markets, introducing new service lines, or adapting to new regulatory requirements, both the company and its engineering team can plan and move without worrying that the software that makes its business go won’t be capable of going along with it.
The Flox Difference
- Nix guarantees, no Nix expertise necessary. Engineers pull an environment and work. The mechanics of Nix are abstracted … unless or until someone needs them.
- The declarative environment is the source of truth. Flox environments are declarative, versionable, distributable. Teams pull and run exactly the same software, every time, anywhere.
- Built-in auditability. Every environment produces a human-readable manifest and an SBOM by default. Compliance is a property of the tooling, not a separate process.
- Change without collateral damage. When requirements or conditions change, detaso can more easily adapt. Flox infrastructure turns change management into changes to declarative manifests, rather than rebuilding images, debugging runtime environments, and other time-wasting activities.


