DOC-001/SOFTWARE ENGINEER/FRONTEND · SYSTEMS · AI/OPEN TO WORK

I build systems that solve product problems.

I'm a Software Engineer who prioritizes user outcomes over clever code. I specialize in bridging the gap between complex technical constraints and business growth.

Next.jsReactNest.jsNode.jsTypeScriptPostgreSQLAI
Portrait photo
§ 01

Selected Impacts

01 / CASE STUDY

Rapid Telehealth MVP Deployment

MVP Launch in 2 Sprints
We Lysn • 2024 - Present
▸ Challenge

A mental health startup needed to move from concept to a functional workplace wellbeing platform within a highly compressed timeframe to validate the market.

▸ Solution

Architected the frontend using Next.js, tRPC, and Supabase. Collaborated in a 'feature trio' to drive rapid discovery and deployment of core telehealth features.

§ Trade-off Ledger
01 · Direct Database-to-Frontend Pipeline
▲ Benefit

Eliminate API layer maintenance for faster shipping

▼ Cost

Coupled deployment cycle

02 · Continuous UI/UX Iteration
▲ Benefit

Allowed technical execution to pivot instantly based on user feedback

▼ Cost

Increased refactoring frequency during early discovery phases

02 / CASE STUDY

Marketplace Performance & Reliability

Optimized Web Vitals
Glints • 2022 - 2024
▸ Challenge

The marketplace platform required better search visibility and extreme reliability to serve millions of career-seekers across Southeast Asia.

▸ Solution

Optimized Next.js performance for SEO and integrated Cypress/Playwright E2E testing into CI/CD to eliminate critical post-deployment regressions.

§ Trade-off Ledger
01 · Design System Maintenance
▲ Benefit

Ensured UI consistency and boosted team shipping velocity

▼ Cost

Ongoing overhead to maintain styled-components library

02 · Automated E2E Testing
▲ Benefit

Reduced critical production bugs, protecting brand trust

▼ Cost

Increased CI/CD pipeline execution time

03 / CASE STUDY

Cloud Migration & UI Unification

+ Client Conversion
Nintex • 2020 - 2022
▸ Challenge

Facilitating complex client transitions from legacy O365 to proprietary cloud systems was a manual bottleneck slowing down revenue growth.

▸ Solution

Engineered a Node.js and NestJS-based migration tool to automate transitions while unifying UI/UX across multiple product suites using React and Angular.

§ Trade-off Ledger
01 · Automated Migration Tooling
▲ Benefit

Lowered barrier for legacy customers to adopt new cloud systems

▼ Cost

High initial development cost to handle legacy data edge-cases

02 · WCAG Accessibility Audit
▲ Benefit

Ensured legal compliance and inclusivity for global users

▼ Cost

Required significant refactoring of legacy component architecture

§ 02

Personal Projects

Fitted In

AI Agents OrchestrationsLLM Streaming

A distributed AI pipeline optimized for zero-latency document synthesis. By fanning out resume processing to parallel agents, the system aggregates a tailored profile in real-time, streaming formatted Markdown directly to a copy-ready workspace for immediate use in Google Docs or Word.

VIEW_CODEBASE
▣ Schematic · Fitted_In/Pipeline
Fig. 01 — Parallel Agent Synthesis
Master Resume
LinkedIn Scrape
▶ Fan-Out Orchestration
Profile Agent
MD_HOOK_GEN
Exp Agent
BULLET_ALIGN
Skills Agent
ATS_FILTER_GEN
◀ Fan-In Streaming
STREAMING_MARKDOWN_CHUNKS…
Final OutputFormatted Workspace
Copy → Docs

My Flix

Affinity EngineStateful UI

A media discovery engine focused on the Reciprocal Loop of user behavior. By weighting sentiment across movie metadata, the system recursively updates a local preference state to re-rank search results and prune discovery feeds, ensuring the interface evolves with the user.

Sentiment Graphing

Links user 'Likes/Dislikes' to specific TMDB entities (Actors, Genres, Directors).

Discovery Pruning

Implicitly removes content from discovery lists based on cumulative negative affinity scores.

VIEW_CODEBASE
▣ Schematic · My_Flix/Affinity_Loop
Fig. 02 — Reciprocal Affinity Feedback Loop
▲ Like
POSITIVE_SENTIMENT
▼ Dislike
NEGATIVE_SENTIMENT
LOCAL_STATE_RECONCILER
Scoring Engine

Weights IDs by sentiment frequency.

Filter Pruning

Excludes content matching negative IDs.

⟲re-ranks & feeds back into discovery
Discovery Feed · Re-Ranked
§ Principles · Engineering Spec

“Build for the users, not the resume.”

I believe that technical debt is often a business decision, not just a coding one. My approach focuses on Maintainable Pragmatism, building what we need today while architecting for tomorrow's scale.

§ 01

Product First: If a feature doesn't move the needle, the code is a liability.

§ 02

Observability: If you can't measure it, you haven't finished the feature.