● Software QA Engineer · Tangerang, Indonesia
Rifaldi Rizqi Pratama.
8+ years breaking things before users do. Now teaching AI agents to draft the tests, so humans can review them.
- playwright
- appium
- katalon
- vitest
- jmeter
- istqb ctfl
Telkom Indonesia
Software Quality Assurance Engineer at PT Telkom Indonesia Tbk, Jakarta. IoT devices, a browser-based VMS, web and mobile products.
QA Automation & AI-Assisted Testing
- Automated regression with Katalon; now moving to Playwright (web) and Appium (mobile) with AI-assisted scripting.
- Consolidating suites for Android, iOS and five web apps (incl. Helix VMS) into one shared repo.
Helix VMS video management for IP cameras
- Tested live and recorded streaming in the browser, including H.265 cameras, across Chrome, Edge, Firefox and Safari on macOS and Windows.
- Wrote SOPs and user guides for setup, daily use and troubleshooting.
Antares Eazy IoT platform
- Functional and integration testing of IoT devices: pairing, remote control, data sync, real-time commands, cloud integration and AI video analytics.
- Kept the Android and iOS apps at 99% crash-free users (Firebase Crashlytics) across 110,000+ users.
- Caught critical pairing, control and sync defects before they reached production.
Process & Leadership
- Standardized test plans, case design and defect triage, cutting each regression cycle by 1–2 days.
- Led test planning and requirement analysis with product and engineering.
- Mentored junior testers to own test design and execution.
- API testing with Postman; JMeter performance tests to find response-time bottlenecks.
Eazy Test Management
The team's test management web app — designed, built and run by me. Test cases, plans, runs, bugs, traceability and dashboards.
Rolling out across 7 projects
- 300+ test cases so far, including Android and iOS runs.
- 800+ Vitest tests behind an 80% coverage gate, plus 33 Playwright E2E tests. Deployed via Jenkins → Docker → k3s.
MCP server & AI test design
- Role-capped MCP server with 32 tools. Agents draft test cases from requirements; QA reviews and approves.
- Audit logs and human-only limits on deletes and overwrites.
Automation runner in development
- Maps Playwright tests to test cases via
@TCtags and posts results straight into test runs.
Multi-agent QA chain
A chain-based agent workflow, so QA reviews AI-drafted test cases instead of writing them from scratch.
Built to fail safely
- Routing logic decides which agent picks up each step.
- Failure handling stops a broken step from poisoning the next one.
- Permission boundaries keep destructive actions with humans.
Things I build after hours
Every one ships with its own tests. Old habits.
Japri
Self-hosted AI assistant for WhatsApp groups: mention alerts, catch-up summaries, reminders, search and image reading. Works with Claude, OpenAI, Gemini, OpenRouter or local models. v0.4.1, one-click installer, 70+ tests.
japri.rifaldirizqi.web.id →Loadsmith
Powerlifting training app: workout logging, Total SBD tracking, rule-based coach and program generation. 70+ Vitest tests.
loadsmith.rifaldirizqi.web.id →RideTalk
Android group voice intercom for motorcycle touring (2–10 riders). PRD, architecture and device QA plan: Bluetooth audio, network handover, 1–4 hour endurance.
kotlin · compose · livekit/webrtcOrdyssa
Sci-fi RPG prototype in Unity. Each milestone validated with manual checklists, 47 automated core checks and per-build smoke tests.
unity · c# · urpSkills
Manual and automated testing, test planning, case design, functional, integration, regression, API and performance testing.
testing tools
ci/cd & devops
development
languages
ai & workflow
methodologies
Certifications & education
- Google CybersecurityProfessional Certificate
- Google Project ManagementProfessional Certificate
- ISTQB CTFLCertified Tester Foundation Level
- Telkom University, BandungBachelor of Informatics Engineering