Carrier Achieves Continuous Quality and World-Class Test Automation Practice with Copado

The Challenge

Manufacturing a Growing Testing Backlog

Carrier is no small operation. As a global enterprise managing Salesforce Sales Cloud, Service Cloud, and Experience Cloud — alongside SAP ERP integrations and mobile applications — the stakes for quality are high and the surface area for testing is vast.

When Deepak Sharma joined Carrier as Salesforce QA Architect, he inherited a testing practice that was entirely manual. Every sprint, every release, every regression cycle was executed by hand — by a team of 10 to 15 people spread across multiple delivery pods.

"It was manual only," recalls Deepak Sharma, Salesforce QA Architect – DSS of Carrier. "It was a cumbersome task — consuming a lot of time to set up the test plan, test strategy, test matrix, and then start execution. Recording everything, categorizing it, checking exit criteria. It was a huge task."

The math was unforgiving. With monthly releases and a regression suite of 150–200 test cases per pod, two QA engineers might be responsible for executing 75 test cases each — in as little as three days. At 10–15 minutes per test case, that was simply impossible to maintain alongside daily responsibilities.

And the backlog only grew. Each release added 10–20% more test cases to the regression suite. What started at 200 cases became 230, then 260, then 290 — with no end in sight. Repetitive manual execution led to team fatigue, missed defects and inconsistent coverage.

The consequences were real. Release windows regularly required team members to stretch beyond working hours. When development ran late — delivering code just days before a release — the team was forced to triage. P3 test cases were skipped. Dormant issues slipped through. Potential risks shipped alongside the product.

The Solution

Building a Low-Code Quality Culture from the Ground Up

When Deepak joined Carrier, his first assignment was clear: evaluate the market and find the right automation tool. He came with deep experience in Selenium and Playwright — but he was looking for something different. "I come from a hardcore automation background — Selenium, years of it," says Deepak. "But I realized that low-code, no-code is something that helps everyone contribute to automation and make it a huge success. Even a person without a technical background can learn the tool and make things work."

After a rigorous proof-of-concept comparing tools across dimensions of maintainability, scalability, cost, and Salesforce-specific capability, Carrier selected Copado Robotic Testing (CRT) based on its crucial functionalities: 

• QWord-based scripting: Human-readable keywords that non-technical QA engineers could learn and use independently
• Zero setup friction: No drivers, no IDE, no executable files. Just a URL and a browser
• Salesforce-native libraries: Dedicated commands like Get ID that abstract complex Salesforce interactions into simple, readable steps
• Parallel execution: The ability to run up to 40 test cases simultaneously, multiplying team capacity without adding headcount
• Mobile testing support: Coverage for Carrier's Go mobile app alongside web and Salesforce environments

"The way Copado has been designed — with human-readable QWords — everyone can understand it regardless of their background,” mentions Sharma. “If you have a coding background, it's icing on the cake. If you don't, you can still learn it and make it work."

Testing Automation Rollout

Carrier partnered with Copado Professional Services for a month-long onboarding and training program. All 10 team members became Copado-certified. From there, Deepak structured the rollout pod by pod — establishing shared keyword libraries, a common folder structure, a Git repository for script versioning, and a team guideline document that stayed open on every engineer's second screen.

The team started small — prioritizing P1 test cases, building confidence with quick wins, then progressively tackling longer, more complex scenarios. Week by week, month by month, the numbers climbed. "Those initial small wins with Copado Robotic Testing were giving us confidence," reflects Deepak. "It was boosting our morale and making us happy that yes — we are progressing."

Scaling Test Automation with Copado Explorer and TestAgent

As the team matured, Carrier adopted Copado Explorer to streamline bug documentation and exploratory testing. Previously, manual testers spent significant time taking screenshots, organizing them in folders, and assembling them into reports one by one.

"Now, whatever you're clicking on — Copado Explorer captures it. Wherever you're typing — it captures the field name and the value. And it puts the screenshot right there," says Deepak. "Explorer is saving us 70–75% of the time we used to spend on documentation."

Carrier also embraced TestAgent, Copado's AI-powered assistant embedded within CRT. For the Carrier team it became an indispensable daily companion.

The Outcome

Success in Numbers

The transformation at Carrier was not incremental — it was exponential. In the first five months of scripting, the team automated 500 test cases. By the end of 2025, they had built 1,200 automated scripts in a single year. Today, 80–90% of Carrier's core Salesforce processes are fully automated, with the remaining 10–20% representing the current sprint — automated in parallel with active development.

But the most striking proof point came from a multi-country regression project. Carrier needed to validate the same 600 test cases across 12 global instances — the same functionality, deployed for different locales. Manually, a single resource would need 15 days per instance. For 12 instances, that was simply not possible.

"Through parallel run, we are running those 600 test cases at the same time for all 12 countries," says Deepak. "All 6,000 test cases are getting executed in just 3 days. There cannot be a better example than this."

Before Copado, the team was forced to prioritize by instance and by test case, accepting that some instances would go untested. Now, every instance, every scenario, every release gets full coverage. What Carrier's team can now execute in 3 days with automation across all 12 instances is more than 2–3x what they could test manually in an entire year with just partial coverage of just a few countries.

With CRT running 40 parallel executions, those same 300-400 test cases that consumed 3 days of human effort now run unattended, freeing each QA engineer to focus entirely on new script development, exploratory testing, and strategic quality work.

The after-hours crunch has been eliminated — automation catches bugs earlier, removing late-stage firefighting from release windows. The bottom line: Copado effectively gave each QA engineer back 6–9 weeks of productive time per year — time that is now being reinvested into building the 1,200+ script library that didn't exist before.

"Automation testing with Copado has significantly reduced the after-hours crunch during release windows," says Deepak. "By catching bugs earlier in the cycle, we've minimized late-stage firefighting. It helps us a lot during debugging and designing new scripts — and it's improved overall team balance and focus."

Team morale has followed. Tools like TestAgent and Explorer have removed the tedious parts of the job such as manual data entry, repetitive execution, screenshot assembly, freeing the team to focus on the high-value detective work that QA engineers actually enjoy. And the time savings with TestAgent is measurable:

• The team saves 40–50% of their time searching documentation or learning new features
• Today there’s a 70%+ reduction in debugging time
• Faster script generation by learning from existing robot file patterns

The Future

From Sprint Planning to System Planning

For Deepak and the Carrier QA team, the journey with Copado is far from over. The next chapter is about moving beyond automation at the test level — and building intelligence into the entire quality lifecycle.

Carrier's near-term goal is to achieve 90%+ automation coverage across all core Salesforce business processes and then expand beyond Salesforce entirely, bringing CRT's capabilities to platforms like Shopify as the business grows.

The team plans to move beyond traditional scripting by integrating Copado Explorer for rapid exploratory sessions and leaning further into TestAgent for a stable, maintainable, and AI-assisted test framework. The goal: shift QA engineers from script writers to strategic quality architects.

Testing Agentforce with CRT

Perhaps the most forward-looking initiative is Carrier's work at the intersection of Agentforce and Copado Robotic Testing. Carrier’s QA team is already building internal Agentforce agents — including one that generates test cases directly from Jira story IDs. And they're using CRT to validate those agents.

"I want to bridge the gap between AI reasoning and action by using CRT to execute the agent's outputs," says Deepak. "Testing an AI agent with CRT is a brand-new frontier for us. I'm excited to dive into the learning curve of how these two powerful tools can work together to close the loop from idea to automated execution."

Using CRT's AI evaluation libraries, which assess agent faithfulness, hallucination rates, and response specificity, Carrier can measure prompt quality and agent reliability with precision. In one example, a precisely engineered prompt scored 90% on CRT's responsiveness scale, versus 70% for a generic prompt. That difference is now a measurable quality signal.

"The more precise you make your prompt, the more favorable the result," says Deepak. "CRT is letting me know when I need to refine my instructions — and that's where it's helping us even to test an agent."

With the ability to test with certainty and implement quality at scale, without mitigating risk, the future is clear: a fully automated, AI-augmented quality pipeline — from requirement to release — where Carrier's QA team spends less time on repetition and more time on the strategic, creative work that drives real business value.