Production Problem Investigation
A clear next step for a recurring production problem.
For agencies, fractional CTOs, and SaaS teams. I trace the failure, test likely causes, and hand your engineers findings they can act on.
Tony St. Pierre16+ years of software development experience
An investigation that fits your team.
Agencies and delivery teams
I work through your technical contact with concise updates and a handoff for your developers. Client-facing communication happens only as agreed.
Fractional CTOs
I provide independent evidence to assign work, advise leadership, or assess a proposed rewrite, with scope and uncertainty explicit.
SaaS CTOs and technical founders
I start with your engineers’ observations and attempted fixes, keeping onboarding focused on the failing workflow and your team’s next decision.
A problem worth investigating
When the same problem keeps coming back.
Complex failures can persist in capable teams. I investigate when repeated fixes haven’t held, senior engineers are tied up, or delivery and client confidence are affected.
My focus is React, Next.js, and TypeScript/JavaScript applications on AWS, including Cognito authentication.
Authentication & Onboarding DiagnosisExamples of suitable problems
- Login, MFA, or account recovery fails for a subset of users.
- Signup or account linking stalls intermittently.
- A workflow works in staging but behaves differently in production.
- A deployment or configuration change introduced an elusive regression.
- Several attempted fixes have failed to resolve the same issue.
What you receive
A written brief your team can act on.
I connect the evidence to your next decision: what to change, what to test, or what to investigate further.
A written investigation brief and practical handoff.
Observed behavior
Symptoms, conditions, and reproduction attempts, including what I could and couldn’t reproduce.
Evidence and reasoning
Relevant evidence, what it supports, and the assumptions behind each conclusion.
Causes investigated
Explanations tested and ruled out, with the evidence for those decisions.
Leading explanation
The explanation best supported by the findings, with confidence and remaining uncertainty.
Next actions
Recommended actions for your engineers and a plan for verifying a repair.
If the cause remains unresolved, I document what was learned, what remains unknown, and the most useful next investigation step.
Start with a short description.
I’ll confirm fit, scope, and timing before we arrange access.
How it works
A defined investigation, from scope to handoff.
One technical contact, a short kickoff, and a closing readout.
Agree scope and access
We define the failing workflow, business impact, existing evidence, access, permitted testing, and scheduled start.
Trace and test
I trace symptoms through relevant code, configuration, and recent changes, then test the explanations supported by the evidence.
Deliver and explain
I deliver the brief and walk your team through the recommendation, confidence, and repair verification plan.
Within the 8-hour cap
I reserve time within the cap for the brief and closing readout. If more investigation is needed, I hand over the findings and next step. Additional work requires an agreed scope and quote.
Emergency incident response, ongoing operational ownership, and unlimited follow-up are outside this package.
Changes by agreement
Repairs are separately scoped after diagnosis. Expanded scope is quoted before additional work begins.
Investigation may include approved testing in an agreed environment. It does not include unapproved production changes.
Your investigator
Tony St. Pierre
I bring 16+ years of software development experience across production web and mobile applications. My work spans application code, identity, integrations, and delivery.
Identity and access
I’ve built secure identity and access controls spanning authentication, MFA, sessions, and account recovery.
Multi-tenant applications
I’ve designed application platforms with clear configuration, feature, and policy boundaries.
Web and mobile delivery
I’ve built production React Native applications and established automated web and mobile delivery with testing and controlled releases.
Before we begin
A few practical questions.
What kinds of problems fit?
One persistent failure or workflow in an existing application, with a team available to act on the findings. React, Next.js, TypeScript/JavaScript, AWS, and Cognito are my focus. A brief description is enough to check fit.
What access is needed?
We agree the minimum access needed: relevant code, configuration, redacted diagnostic evidence, and an environment for approved testing. Detailed evidence follows the scope and access agreement. Start with a summary by email.
What if the root cause remains uncertain?
I don’t guarantee root-cause discovery. I deliver what was learned, explanations tested, remaining uncertainty, and the most useful next investigation step. Any further work needs a separate agreement.
Are fixes included?
Repairs are separately scoped after diagnosis. Your brief includes recommendations and a repair verification plan. Expanded scope is quoted before additional work begins.
Can you work through my agency?
Yes. We agree communication and handoff arrangements at kickoff. I work through your technical contact and communicate with your client only as agreed.
When does the delivery window begin?
Delivery takes 5 business days. The window begins after scope is agreed, necessary access and evidence are available, and the scheduled engagement begins. The window is elapsed time; the effort cap is 8 total hours, including documentation and calls.
Start with a summary
Tell me what keeps happening.
A few lines are enough. I’ll confirm fit, scope, and timing before we arrange access.
In your email, include:
- Role and company
- Application stack (if known)
- Symptoms and attempted fixes
- Business impact
- Desired timing
Please omit credentials, tokens, personal data, and raw production logs. Detailed technical evidence can be exchanged after scope and access arrangements are agreed.