Probe Droid QA Automation System

Home / Projects / Probe Droid QA Automation System
Hims & Hers Health

Hims & Hers Health

2025

Overview

Developed Probe Droid as a multi-channel campaign QA system that made repetitive validation faster, more consistent, and easier to inspect before launch.

The Problem

Campaigns often combine audience rules, events, attributes, content, links, suppression logic, and channel-specific settings. Reviewing each part manually creates a slow final checkpoint and makes defects easy to miss, especially when the same campaign must be validated across email, SMS, push, social, paid, and in-app surfaces.

The goal was not simply to automate a checklist. It was to create a repeatable validation layer that could identify risk early, show the reason for each flag, and leave the campaign owner with a clear path to resolution.

Architecture & QA Loop

Probe Droid follows the campaign from request through launch readiness. It evaluates structured inputs, checks cross-system relationships, and returns an actionable report rather than a single pass/fail result.

  1. Collect — gather campaign metadata, channel configuration, audience definitions, content, and timing.
  2. Validate — check required fields, data types, naming, links, event mappings, and platform constraints.
  3. Trace — verify that events and attributes can travel from source systems to targeting and activation.
  4. Review — test personalization, eligibility, suppression, and channel-specific behavior.
  5. Report — group issues by severity and explain what needs to change.
  6. Monitor — compare launch behavior with expectations and surface anomalies for follow-up.

Key Decisions & Tradeoffs

  • Put validation before launch, where fixes are cheaper and safer.
  • Treat suppression and eligibility as quality requirements, not optional checks.
  • Return explainable flags so operators can fix the source rather than work around the symptom.
  • Use a shared framework across channels while allowing channel-specific rules.
  • Keep human approval for business judgment and exceptions that cannot be safely encoded.

Key Highlights

  • 90% reduction in campaign validation time (1 hour → 2 minutes)
  • Multi-channel support (paid, social, email, SMS, push, app)
  • 132x productivity gain in content creation
  • 67% reduction in campaign error rates (6 flags → 2 flags per ticket)

Results & Evidence

The public operating measures for this project report a reduction in validation time from about one hour to about two minutes, along with fewer recurring campaign flags. Those figures describe the measured workflow improvement; they do not imply that every campaign or channel receives identical savings.

Company scale, customer data, internal identifiers, and implementation-specific details are intentionally omitted. Results should be interpreted in the context of the workflow and measurement method.

Lessons for Other Teams

  • Cross-channel QA should test the data path, not only the final message.
  • Automated checks create more value when their output is understandable to non-engineers.
  • Launch monitoring belongs in the same quality system as pre-launch validation.
  • Standardized campaign metadata is the foundation for reliable automation.

Technologies

Automation QA Systems Multi-channel Marketing Process Optimization