All work

Mosaic

Stabilizing a production POS without pausing product delivery.

Reliability and modernization work on a production Flutter point-of-sale application. I investigated recurring ANRs, crashes, runtime errors, and slow interactions; replaced several ANR-prone WebView paths with native Flutter flows; and added Firebase instrumentation so production evidence could guide the next improvement.

Role
Product Engineer · Shuru Technologies
Platform
Flutter · Dart · Android runtime · Firebase
Status
Production POS modernization
Period
Oct 2023-Sep 2025

Project evidence

Production POS
production application
WebView → Flutter
several ANR-prone flows rebuilt
ANRs + crashes
recurring failure modes investigated
Firebase
production instrumentation and analytics

The first obligation

Make the workflows people depend on reliable again.

Mosaic was an active production POS application with recurring ANRs, crashes, runtime errors, and slow interactions across critical workflows.

I profiled and debugged those failures while continuing normal product delivery. The priority was to remove user-facing instability without assuming a rewrite was available or desirable.

Targeted modernization

Replace the paths that repeatedly failed, not the entire product.

Several recurring ANRs came through WebView-based paths. I rebuilt those workflows as native Flutter experiences and progressively refactored brittle legacy code around them.

The approach improved application structure in bounded steps and avoided disrupting ongoing production delivery.

System map

A production reliability loop

Instrumentation and user evidence identify risky paths; profiling narrows the cause; bounded native replacements ship; production evidence validates the next step.

  1. 01Production signals

    Firebase · runtime errors · user feedback

  2. 02Profile and debug

    ANRs · crashes · slow interactions

  3. 03Isolate the risky path

    WebView and brittle legacy boundaries

  4. 04Native Flutter flow

    Bounded replacement and refactor

  5. 05Production release

    Continue product delivery

  6. 06Observe the result

    Use real-world evidence for the next cycle

  • Production signals connects to Profile and debug.
  • Profile and debug connects to Isolate the risky path.
  • Isolate the risky path connects to Native Flutter flow.
  • Native Flutter flow connects to Production release.
  • Production release connects to Observe the result.
  • Observe the result connects to Production signals.

Close the loop

Instrumentation turns isolated reports into an engineering signal.

I added Firebase production instrumentation and analytics, then combined real-world feedback with profiling and debugging evidence.

That loop made performance, responsiveness, and reliability work observable rather than dependent on reproducing every failure locally.

Proprietary production code

About the source code.

The source code is proprietary to Shuru Technologies. I can discuss my responsibilities, profiling and debugging approach, technical decisions, trade-offs, and outcomes without disclosing confidential material.