How First Orion Achieved 40% Faster Releases with Programmable Device Cloud

At a glance

• Up to 40% faster cycle time for mobile app and telecom solutions
• 35% of engineering and QA capacity freed for higher-value product work
• More than 10% reduction on physical device spend alone
• Zero hands-on device management and maintenance

‍

About First Orion

First Orion operates at one of the most consequential moments in modern communications: when a phone rings or a message arrives. The company shields wireless subscribers and business users from scammers and bad actors. It also powers branded communications by helping businesses present a verified, trusted identity when reaching their customers through calls and messages.

Both facets require the same underlying guarantee: that what a consumer sees on their screen is correct, consistent, and trustworthy. That guarantee has to hold across every enterprise customer, every device model, every OS version, and every network condition. For a company whose product is the phone call experience, testing is a business-critical function that underpins how First Orion delivers predictability to the telecommunications customers and enterprise clients that depend on it.

The challenge: The gap between test environment and production reality

Mark Himelfarb, First Orion's CTO, describes the testing challenge as analogous to an iceberg. You need to see what's happening on the device and understand the user experience — but the real complexity lies beneath the surface.

Testing a branded call isn't simply confirming that a logo appears. It requires understanding the full telecommunications infrastructure — every hop between the originating network, the enterprise customer, and the receiving device — to know why an experience is rendering a certain way. That diagnostic depth is what separates a company that can guarantee outcomes from one that can only observe them.

Before Sauce Labs, First Orion's approach was entirely manual.

"Piles of devices on desks. We called it 'musical phones.' People on the West Coast, in central Arkansas, on the East Coast, and abroad. Everybody needed to be able to do these tests, and you were either having to maintain piles of phones, update them, charge them, make sure they're in the right configuration, and make sure the right SIMs are in them. And then you had to maintain duplicates," said Himelfarb.

The overhead was more than a logistical inconvenience. Himelfarb estimates that, in some cases, up to 10% of team members' time was consumed simply getting to a state where a single test could be run. With engineers and QA staff distributed across time zones, device configurations drifted. Automation that worked on one machine failed on another.

The forcing function became clear as First Orion's business grew: telecom customers and enterprise clients needed confidence that branded calls and messages would render correctly across an ever-expanding matrix of devices and OS versions. Manual testing at a desk couldn't deliver that level of assurance at scale. Himelfarb’s assessment was plain: "Ineffective, highly labor-intensive, and most importantly, absolutely unscalable."

The decision: Non-negotiable requirements: the bar other vendors couldn’t clear

First Orion needed real devices with real SIM cards on real U.S. carriers. The team needed the ability to receive and originate actual phone calls and SMS messages. And it needed the solution to be programmatically controllable at every layer of the stack.

Because First Orion's testing workflow had to integrate signals from across the telecom infrastructure (device logs, network behavior, inbound and outbound call events), the testing platform had to offer an open, extensible API. A closed portal with a fixed set of operations wouldn’t work.

"Some vendors provide programmatic interfaces, but they're very severely limited — you can only do this but not that, you cannot export or import data, you cannot manage traffic or emulate various conditions. We did not find those things in other vendors," said Himelfarb.

Most software testing customers are validating an app's UI or user flow. First Orion needed to test the cellular communication layer underneath the UI — a requirement that demanded a vendor capable of engaging with unusual needs, not just tolerating them.

Sauce Labs cleared both bars. Its Real Device Cloud offered genuine device diversity across Apple, Google, and other OEM hardware with live SIM cards on U.S. carriers. The recently released Real Device Access API gave First Orion the programmatic control and depth that Sauce Labs’ competitors couldn’t match.

Implementation: Where real devices meet programmable infrastructure

First Orion didn't simply plug into Sauce Labs and run off-the-shelf test suites. It used the Real Device Access API as a foundation for custom internal tooling — infrastructure that wouldn't have been possible on a more constrained platform.

Some benefits of using the Real Device Access API include:

• Real-time issue visibility: Automated test sessions don't expose live screen state by default. First Orion built tooling on top of the Real Device API that captures and mirrors everything happening on a device's screen during an automated session. Support and QA teams can monitor experiences in real time and replay sessions to pinpoint failures, transforming reactive triage into a proactive feedback loop.
• Live log streaming with a unified view of signals: First Orion's most sophisticated custom solution uses Sauce Labs' log streaming capabilities to capture device logs in real time, then connects those logs with signals from other parts of the telecom infrastructure. The result is a unified view of what's happening at every layer — network, carrier, and device — simultaneously. That single source of truth is what makes root cause analysis possible at speed.
• Persistent sessions: The Real Device Access API's persistent session model allows First Orion to maintain a continuous connection to a device for up to 18 hours—all through HTTP requests. This eliminates repetitive initialization cycles and setup and teardown, particularly valuable when testing real-time telecom events.

Adoption expanded quickly beyond engineering. What began as a narrowly scoped use case grew to include QA, product, support, and eventually sales and customer success teams, who began using the Sauce Labs platform for product demonstrations and proof-of-concept walkthroughs with prospects.

Results: What release confidence at scale looks like

The outcomes span direct cost and capacity recovery, improved cycle time, reduced overhead, and a strategic shift in how First Orion builds products.

‍Capacity reclaimed for higher-value work: 35% of combined engineering and QA capacity — previously consumed by device setup, configuration management, and maintaining physical inventories across distributed offices — is now redirected to product development, architecture work, test design, and quality analysis. Engineers are deepening product work while QA teams are focused on test strategy, not logistics.

‍Faster and more predictable release cycles: Automated parallel testing across device groups, OEMs, and OS versions delivered a baseline 20% improvement in release cycle time across products. For mobile application and telecom solutions specifically, the improvement reached as high as 40%. First Orion now ships with confidence, not uncertainty.

‍Cost optimized and contained: The team eliminated physical device inventory across distributed offices, removing the purchase, maintenance, and configuration costs that had scaled with every new hire and every new location. Direct hardware savings came in at 10%+ on physical devices alone — and unlike the old model, those costs no longer compound as the company grows.

Innovation and product development unlocked: Engineering teams used the freed-up time to deepen product work and pursue new feature development. UI/UX engineers redirected effort toward refining existing products and prototyping new ones. Product teams adopted the platform for proof-of-concept development, using live device visualization to demonstrate concepts earlier in the build cycle.

First Orion is also running AI models against signals captured through Sauce Labs devices and Real Device Access API by analyzing telemetry data to surface insights and automate decision-making faster than manual review allows. The platform has become a data source, a development tool, and an innovation accelerator, not just a testing environment.

"The capability we gained with Sauce Labs to run tests at a very large scale, across multiple device groups, multiple operating system versions, and OEMs has been truly fundamental. We would never be able to do this manually at this level. That's a huge win," said Himelfarb.

Strategic partner for First Orion's product & AI roadmap

Sauce Labs' impact at First Orion goes beyond any individual metric or team. When new product concepts emerge, the platform is now assumed to be part of the architecture.

Himelfarb recently talked with First Orion's website team, which had encountered a device-specific rendering issue after a launch. The team hadn't considered using Sauce Labs for web testing. His response: "Did you not know about Sauce Labs?" They do now. The platform that began in a single engineering team is now a shared organizational capability.

First Orion is also evaluating Sauce Labs' role in its AI strategy, including the potential integration of an MCP server into its development workflow. The combination of Sauce Labs device telemetry and AI-powered analysis is foundational to how the company intends to build and validate next-generation communication products.‍

"We've now got product ideas that are going to rely very heavily on Sauce Labs. We're actually counting on it being part of our product strategy, which I think is really incredible," said Himelfarb.

For technology leaders navigating a similar decision, Himelfarb's advice is direct: “Take the dive, do it, explore Sauce Labs, and spend the time understanding the level of sophistication that they bring to the table. I would absolutely recommend them as a company, recommend them as a vendor, and most importantly, recommend them as a technology partner."