How Attorney Share Scaled Their Platform Without Sacrificing Quality
“Our QA engineer from Distillery is an MVP member of our team. They’ve developed deep knowledge of our product and our business. They’re a valuable thought partner when it comes to designing new features and experiences on the platform.”
— George Durzi, CTO, Attorney Share
There’s a quiet shift happening in early-stage tech. With tighter budgets, compressed timelines, and pressure to ship fast, more CTOs are making the same bet: that a great dev team can carry the weight of testing without a dedicated QA.
Sometimes that works…until it doesn’t.
At Distillery, we work with lean teams all the time. And we’ve seen firsthand how powerful a single QA can be when they’re brought in early, given the right tools, and trusted to take ownership of quality from day one.
So what did this actually look like in practice? Let’s talk about Attorney Share.
The Setup: Lean Team, Complex Platform
Attorney Share is a legal tech startup building a modern case referral platform. Their engineering team is small, just 13 people, including 4 frontend developers, 4 backend developers, a PM, a BA, and one QA engineer.
The product itself is anything but simple. It’s multi-tenant, serves different legal organizations, and includes workflows that require validation across orgs, users, and data layers.
Despite all that complexity, the team hasn’t had a major production issue in over a year and a half. Releases are smooth. Stakeholders are confident. And their QA function is fully led by one Distillery engineer.
How It Worked
Initially, the approach was lean and manual, centered on exploratory testing and high-priority risk coverage. Documentation was kept light but purposeful, using mind maps, diagrams, and checklists instead of bulky test plans. This helped the team stay agile while the product was still evolving.
As the platform matured and release velocity became more important, automation was introduced strategically. The goal wasn’t just to automate for speed; it was to build a sustainable framework that could evolve alongside the product. That’s where AI came in.
A custom framework was developed in TypeScript, using Playwright for UI testing and Vitest for API validation. The results were fast and effective: test runs clocked in under 30 seconds and covered 70% of core features. But the real impact came from how AI was integrated into the process.
Instead of relying on AI for small tasks, the QA workflow was restructured around a more advanced toolset. Tools like Claude Code and MCPs became core to framework development and maintenance. These tools made it possible to generate complex test scripts directly from requirements, refactor code dynamically, and adapt the framework as the product evolved – all with significantly less manual effort.
The result was more than just a functional automation suite. It was a flexible, intelligent system that expanded what one QA engineer could realistically accomplish. Exploratory testing remained essential for catching nuanced edge cases and user behaviors, but AI took on the heavy lifting, allowing the QA to focus on high-impact, strategic work.
The Results
Here’s what one embedded QA achieved in two years:
- 228 defects identified – 25 critical, 45 high severity
- 70% automation coverage for core product features
- Full-stack test coverage across API, DB, and UI
- Test automation execution under 30 seconds
- Custom documentation and framework development
- No major production issues for 18+ months
- Deep team trust and high stakeholder confidence
This wasn’t about throwing bodies at the problem. It was about embedding the right person, with the right mindset, at the right time.
And that mindset? It’s something we teach and invest in at Distillery.
What Makes Distillery QA Different
At Distillery, our testers are trained to do more than write and run test cases. They’re encouraged to become business experts, to ask uncomfortable questions, and to take ownership of the product’s health, not just its correctness.
We run mandatory training sessions for all our QA talent, led by respected voices in the testing world. Topics include strategy, systems thinking, and mental models, because testing isn’t just execution, it’s insight.
That’s how we build QA professionals who can walk into a lean team and make an outsized impact.
Why This Matters for Other Teams
The point of this story isn’t that every startup should hire a QA engineer tomorrow. It’s that there’s a middle ground between no QA and a full department, and when done right, that middle ground is incredibly effective.
If your team is small, your roadmap is aggressive, and you’re trying to keep burn low, this model is worth considering.
A single, embedded QA, with a smart test strategy and support from automation and AI, can give you faster feedback loops, fewer fire drills, and more confidence in what you ship.
And you don’t have to take our word for it.
Want to learn how this model could work for your team? Reach out to us for a free 30-minute consultation.