Ambiguous failures usually aren’t ambiguous.
The team just doesn’t have enough information yet. My work in root cause analysis begins by making the unknown visible—then giving people a shared way to act on what we learn.
A shared RCA process built around recurring defects, ownership, and follow-through.
18+
engineering teams coordinated through one process.
Adopted.
Leadership support and formal adoption within Software Engineering.
I start by making the unknown visible.
QA often arrives after the environment has already taken shape. The first responsibility is to learn what is actually there, document it clearly, and build the shared context the investigation has been missing.
Establish the source of truth
I track down the people, behavior, documentation, and code history that explain how the product is supposed to work—and where those accounts disagree.
Reconstruct how the product behaves
I learn the system inside and out instead of treating the latest symptom as the whole problem. That creates a living QA record the team can keep using.
Question accepted assumptions
Teams move fast, and familiar decisions can become invisible. An independent QA perspective makes room to ask why something works this way and whether it still should.
Connect evidence across teams
A recurring failure rarely respects organizational boundaries. I bring the right people and evidence together so the investigation has one shared context.
Turn findings into institutional knowledge
The investigation should outlive the incident. What the team learns becomes clearer documentation, ownership, follow-through, and a stronger quality practice.
What good QA leaves behind
QA should become part of the team’s institutional memory.
AI helps me
- Organize logs and documentation
- Connect evidence across large systems
- Surface patterns and missing information
- Preserve what the investigation teaches us
Human judgment decides
- Whether an outcome makes sense
- Whether it feels right to the person using it
- Whether a response is responsible and ethical
- What visual behavior means in spatial products
AI can help carry the weight of investigation, synthesis, and prevention. The human element still matters because the product is made for humans—especially in spaces like AR and VR, where what technically appears and what visually makes sense are not always the same thing.
A working diagnostic
Bring the failure into focus.
Describe what happened. The RCA Lab will organize the evidence, test assumptions with the 5 Whys, and return a practical investigation plan. It is a first pass—not a substitute for learning the system.
See the process take shape.
This interactive map shows how focused testing, engineering context, and shared evidence can move a team from a recurring symptom toward accountable follow-through.
Explore the process map
Bring me the problem your team can’t clearly explain.
The diagram shows where I begin. The deeper work is learning your system, reconstructing its history, connecting its evidence, and building a quality practice around what we find.
Let’s find out what information is missing.
Start a conversation