SOFTWARE ENGINEER  /  SYSTEMS BUILDER

I build systemsthat make softwaremore reliable.

I'm Karyo. I work at the intersection of software engineering, AI-driven automation, and quality infrastructure, turning complex delivery problems into dependable systems.
Explore selected work The person behind it ↓
PRODUCT ENGINEERINGINTELLIGENT AUTOMATIONRELEASE RELIABILITY
01 / SELECTED WORK

Built with purpose.
Engineered to last.

Three initiatives. One belief: quality should be engineered into the system, not inspected at the end.

01   /   FLAGSHIP PROJECTVALIDATED MVP

VisualGuard.

A visual regression platform for comparing environments, reviewing UI differences, and building confidence before release.

PlaywrightCloudflareD1 / R2
◈ VisualGuard.
WORKSPACE / OVERVIEW
PROJECT / LAW LIBRARY / VISUAL REVIEW

Comparison overview

Pages reviewed21
Identical20
Review items01
Environment comparisonVIEW DIFFERENCES ↗

What I built

VisualGuard combines a Cloudflare-based application, D1 for structured records, R2 for image artifacts, and a Playwright-powered runner. It supports production and staging baselines, screenshot comparisons, and human review of flagged differences. The interface pictured here is an illustrative concept, not a screenshot of the production dashboard.

02   /   AI SYSTEMSIN DEVELOPMENT

AI Triage
Agent.

From failed tests to verified findings. An AI-assisted triage workflow designed to surface the signal, collect evidence, and reduce manual investigation.

AI AgentsJenkinsNode.js
01 / INPUTTest failure
detected
02 / REASONEvidence-led
AI diagnosis
03 / VERIFYReviewable
finding
How it works

Designed as deterministic pre-triage, AI-assisted investigation, then verification before reporting. Jenkins orchestration, evidence archival, and Teams/Jira reporting are part of the planned workflow. This is an in-development system, not a finished autonomous service.

03   /   PRODUCT VISIONCONCEPT

Release
Guard.

A release intelligence concept bringing change impact, test evidence, and operational risk into a clearer decision about deployment readiness.

Release EngineeringCI/CDAI
RELEASE
SIGNALS
CHANGE IMPACT↗
TEST EVIDENCE↗
RISK REVIEW↗
The idea

ReleaseGuard is envisioned as an engineering decision layer that combines automated validation, impact analysis, policy checks, and human judgment. It is currently a product concept and is intentionally not presented as production software.

02 / HOW I THINK

Less ceremony.
Better systems.

I work across development, infrastructure, and quality engineering, connecting the pieces that are often treated as separate problems.

01 /

Engineer for reliability.

Make failure modes visible, behavior testable, and production changes easier to trust.

02 /

Automate with evidence.

Use automation and AI where results can be inspected, reproduced, and verified.

03 /

Build for the long term.

Prefer clear architecture, operational simplicity, and systems that teams can maintain.

03 / THE PERSON BEHIND THE SYSTEMS

An engineer at heart.
A builder by nature.

I’m a software and quality engineering professional with hands-on experience building test platforms, automation frameworks, CI/CD pipelines, and cloud-backed developer tools.

These days, I’m focused on the next layer of software delivery: intelligent systems that help teams understand what changed, what failed, and whether a release is truly ready.

TypeScriptPlaywrightCypressNode.jsCloudflareDockerJenkins
04 / WHAT'S NEXT

Let's build
what's next.

Interested in building reliable software, developer platforms, or AI-assisted engineering workflows? Let's connect.