Own the Product, Not the Code: A Story of Collaboration and Engineering Maturity

In any engineering project, bugs are inevitable, they are simply part of the journey. But what happens when the pressure to deliver outweighs formal QA and operational frameworks? Debugging then becomes much more than a technical task; it becomes a reflection of an organization’s strategic and cultural maturity. The PX project was exactly that kind of test.

Context: Precision by Design

The PX project followed a fast-paced delivery rhythm, one release per week with clear feedback loops and transparent release notes that simplified change management among stakeholders. We embraced a NoQA culture, where quality is everyone’s responsibility rather than a separate team’s. Peer reviews, cross-testing, and Agile/Scrum rituals ensured operational discipline.

Yet, this efficiency came with a caveat: success often depended on individual heroics rather than repeatable, scalable systems. The organization was still in a “Heroic” maturity phase, characterized by ad hoc processes, undocumented knowledge, and unpredictable outcomes.

Challenge: Delivery at the Forefront

From day one, the goal was simple: deliver quickly and well. Testing was pragmatic and focused, prioritizing core functionalities under tight deadlines. This approach optimized speed without neglecting quality, but it leaned heavily on reactive, incident-driven problem-solving.

As the project approached acceptance testing, the need for resilience over mere functionality became clear. We needed to stress-test the system under real-world conditions to prepare for launch.

How we made it: From Chaos to Clarity

At this pivotal moment, the Head of Engineering gave a game-changing directive:

"Test the app like an end user. Forget the code you know. Just explore."

With no scripts or QA tooling in place, the team ran a full weekend of stress testing. The mindset shifted from "debug what you wrote" to "experience what a user would endure." That unstructured, curiosity-driven approach uncovered several bugs, many of which had remained invisible to traditional dev workflows. Initially chaotic, this process became a catalyst for transformation.

From Heroic to Proactive

This wasn’t just about fixing bugs, it was about rewriting the team’s operating model.

We abandoned the 'hero culture', the reliance on a few senior engineers pulling last-minute saves and instead invested in collective ownership of quality. We started working as one team.  Frontend, backend, infra everyone was in the same (virtual) room, reproducing issues live, sharing logs, and solving problems in real time. Process replaced improvisation. Every bug was logged, and tracked. We introduced a light but effective ticketing process that made issue ownership and resolution traceable. We began to ask “why did this bug happen? Instead of hotfixes, we diagnosed the origin of each issue using structured logs and set preventive measures in motion. Cultural reframing of bugs. Errors were no longer viewed as failures but as data points, each one contributing to our growing knowledge base and fueling process improvement.

Logs became our compass in our emerging engineering culture: “Code without logs is useless.”

No bug was considered fixed unless it could be reproduced, traced, and explained. This marked a true maturity leap: from a reactive firefighting mode to a managed, proactive posture, where every error was an opportunity to improve the system and the team.

Impact: Building Resilience Together

By the end of this high-pressure sprint, we didn’t just ship a working app, we laid the foundation for an engineering culture built on resilience and shared learning.

• The system became stronger through real-world pressure, like an immune system building antibodies after exposure.
• Team dynamics evolved. Engineers started quoting logs in standups, tracking patterns, and suggesting systemic fixes without being prompted.
• New habits were created. We continued testing together. We made sure logs were clear and helpful. After fixing big issues, we started doing short reviews to remember what we learned.

Most importantly, we began treating incidents as collective learning moments rather than isolated crises. This cultural shift marked the beginning of a more strategic, proactive engineering mindset, unlocking the next phase of evolution: smarter tooling, AI-assisted observability, and eventually, autonomous error detection.

What’s Next?

Our next story shifts to a new kind of challenge: building a custom modeling library for actuaries. No dashboards, no black boxes, just code, transparency, and collaboration.

What does it take to replace a market-leading tool with something tailored, open, and equally powerful? Let’s find out.