Back to Portfolio
Test Automation · QA Strategy · CI/CD Integration

QualityGate — Test Automation Overhaul for Northbeam Analytics

Replaced a slow, unreliable manual QA process with a maintainable automated test suite for a fast-growing analytics SaaS platform — cutting regression testing from 4 days to 3 hours.

Playwright GitHub Actions API Testing Visual Regression Test Strategy
4 Days → 3 Hrs
Regression Test Time
70%
Fewer Post-Release Defects
3x
Faster Release Cadence
3 Months
Delivered
Delivered
Test Automation
4 Days → 3 Hrs
Regression Test Time
70%
Fewer Post-Release Defects
Client
Northbeam Analytics
Industry
Software / SaaS
Location
Toronto, Ontario, Canada
Duration
3 Months · 2025
Project Overview

About This Project

Northbeam Analytics was releasing new features every two weeks, but each release required four full days of manual regression testing across their growing platform — testing that was inconsistent between QA staff, frequently missed edge cases, and was the single biggest bottleneck in their release cycle. Two significant bugs had already reached production despite the manual process.

Digivance designed a risk-based test automation strategy, then built a maintainable Playwright-based automated suite covering the platform's highest-risk user flows and API surface, fully integrated into their GitHub Actions CI/CD pipeline. Regression testing that took four days now completes in under three hours, automatically, on every release candidate.

Playwright
GitHub Actions
API Testing (Postman/Newman)
Visual Regression Testing
Risk-Based Test Strategy
4 Days → 3 Hrs
Full regression suite execution time
70%
Reduction in defects reaching production
85%
Of critical user flows now covered by automation
3x
Increase in safe release frequency
The Problem

Challenges We Solved

4-Day Manual Regression Cycle

Every two-week release required four full days of manual regression testing, consistently becoming the critical-path bottleneck that delayed releases or forced QA to work weekends.

Inconsistent Manual Test Coverage

Different QA staff tested different things depending on who was assigned that week, meaning coverage was inconsistent and important edge cases were regularly missed.

Bugs Reaching Production

Two significant defects had reached production in the six months before Digivance was engaged, each requiring an emergency hotfix and damaging customer trust in a still-growing product.

No CI/CD Test Integration

Testing happened entirely outside the deployment pipeline as a separate manual phase, meaning there was no automated quality gate before code reached production.

Testing Couldn't Keep Pace With Growth

As the platform's feature surface grew, the manual testing burden grew faster than the QA team could scale headcount to match, threatening to slow the company's release cadence permanently.

No Documented Test Strategy

There was no risk-based prioritisation of what to test — QA effort was spread evenly across all features regardless of actual business risk or usage frequency.

Our Approach

How We Solved It

Risk-Based Test Strategy First

Before writing any automation, we mapped Northbeam's platform by business risk and usage frequency, identifying the roughly 20% of user flows responsible for 80% of real customer impact — and focused automation there first.

Maintainable Playwright Framework

Built a Playwright-based automation framework using stable selector strategies and proper abstraction patterns, specifically designed to avoid the flaky, high-maintenance automation that causes teams to abandon test suites.

API-Layer Test Coverage

Added API-level testing beneath the UI layer for faster, more reliable coverage of business logic — catching backend regressions in seconds rather than waiting for slower, more brittle UI tests.

Visual Regression for UI-Critical Flows

Implemented visual regression testing for the platform's most visually complex dashboards, catching unintended layout and rendering issues that functional tests alone would miss.

Full CI/CD Pipeline Integration

Integrated the automated suite directly into GitHub Actions, so every pull request triggers relevant tests automatically — creating a genuine quality gate before code can reach production.

QA Team Enablement

Trained Northbeam's QA team to write and maintain new automated tests themselves going forward, ensuring the automation investment keeps growing after our engagement rather than stagnating.

Results

The Outcomes

Month 1 — Strategy & Framework
Risk-Based Test Plan & Framework Built

Completed risk-based test strategy and stood up the core Playwright automation framework, with the first 15 highest-risk user flows automated and passing reliably in CI.

Month 2 — Coverage Expansion
85% of Critical Flows Automated

Expanded automated coverage to 85% of identified critical user flows, plus API-layer and visual regression testing for the platform's core dashboards.

Month 3 — Full CI/CD Integration
4 Days to 3 Hours, Zero Flaky-Test Abandonment

Full suite integrated into GitHub Actions as a required release gate. Regression testing time dropped from 4 days to 3 hours, with post-release defects down 70% in the first quarter of operation.

4 Days to 3 Hours. 70% Fewer Production Bugs.
QualityGate gave Northbeam Analytics a maintainable automated safety net that scales with their release cadence instead of becoming a bottleneck to it.
★★★★★
"Our QA team used to spend every release week buried in manual regression testing, and bugs still slipped through. Digivance built us something we can actually maintain ourselves — three hours instead of four days, and we've shipped three times as often since, with fewer incidents, not more."
AD
Aisha Diallo
Director of Engineering, Northbeam Analytics

Stop Letting Testing Slow You Down

Let's build a test automation suite that actually scales with your release cadence.

Back to All Portfolio Case Studies