Cedrick Cantero

Software that has to be right, built by Cedrick Cantero.

Senior Software Engineer · Full Stack · currently at Atmail Inc. · Cebu, PH → Remote

Seven years building systems in regulated and transactional domains: property settlement, ISP email activation and billing, ERP integration, and payments. Systems where a wrong figure becomes a legal or financial mistake.

View workDownload resumeGitHubLinkedInEmail
full-stack experience
7+ yrs
full-stack experience
initial load after SSR rewrite
−40%
initial load after SSR rewrite
E2E coverage on billing flows
~95%
E2E coverage on billing flows
users on shipped dashboards
10k+
users on shipped dashboards

About

Cedrick Cantero, software engineer, professional headshot

Backend and platform work includes server-side validation of land title and mortgage records for Australia's national property settlement platform, NetSuite SuiteScript, and idempotency design spanning application database and payment gateway.

Front-end depth across React, Remix SSR, and React Native, including a monorepo migration and an SSR rewrite that cut initial load time 40%. Shipped a fine-tuned LLM in a production data pipeline.

Experience

Atmail Inc. · Software Engineer

FEB 2024 - PRESENT
  • Own the 3-step activation and billing enrollment flow (KYC verification, plan selection, and Chargebee payment capture), with duplicate enrollment made impossible by design and ~95% Playwright E2E coverage.
  • Migrated a client-rendered React app to Remix SSR, collapsing an 8–10 call waterfall into 3–5 parallel loaders: initial load down 40% on Lighthouse and real-user monitoring.
  • Remediated a stored XSS in the admin HTML preview, sanitizing the render path without breaking the feature.
+ 5 MORE
  • Engineered the My Account dashboard, a customer-facing system surfacing quota, storage allocation, plan type, and subscription lifecycle state, serving 10,000+ users.
  • Architected a TurboRepo monorepo across 5+ packages, improving CI/CD pipeline efficiency by 30% and reducing inter-package build dependencies.
  • Profiled and resolved system bottlenecks using Lighthouse and React Profiler, cutting average load time by 1.8s and reducing rendering latency by 25%.
  • Led a system modernization initiative, migrating legacy class-based React architecture to hooks-based patterns, reducing component coupling and improving long-term maintainability.
  • Introduced AI-assisted engineering workflows using Windsurf and Claude Code, authoring structured codebase briefing docs that enforced consistent patterns and accelerated feature delivery across the team.

REACT · REMIX · TYPESCRIPT · CHARGEBEE · TURBOREPO · PLAYWRIGHT

Accenture Inc. · Software Engineer (PEXA)

JUN 2021 - FEB 2024
  • Built Java modules validating land-title and mortgage records for Australia's national property settlement platform: correctness work on legally binding transactions.
  • Engineered 40 of 68 modules in a live JSP-to-React migration, setting the state-management and validation standards adopted team-wide.
  • Eliminated a production crash class with response validation before render and error boundaries to contain failures.
+ 2 MORE
  • Designed and maintained multi-environment CI/CD pipelines (Git, Bamboo, Jenkins), improving release reliability and deployment velocity across development, staging, and production.
  • Mentored junior engineers in software engineering best practices; conducted structured peer code reviews to enforce system quality and consistency.

JAVA · REACT · JSP · BAMBOO · JENKINS

Xlr8 Ventures Inc. · Software Engineer

NOV 2020 - APR 2021
  • Built enrollment, gradebook, and attendance modules for a school-management SaaS, with a three-tier role-based access scheme scoping every view.
  • One of two engineers on the team; cut build times ~20% with Webpack code splitting.
+ 1 MORE
  • Implemented independently against written specification, with implementation approach (structure, state management, component patterns) at my discretion.

REACT · WEBPACK · RBAC

Excelym IT Solutions · Software Engineer

JUL 2018 - OCT 2020
  • Wrote server-side SuiteScript inside NetSuite ERP: validation, transforms, and lifecycle scripts on live order and inventory data.
  • Designed retry gated on terminal failure state, making duplicate ERP orders structurally impossible; built the operator console that classified failures by whether a retry could ever succeed.
+ 2 MORE
  • Built paginated, filterable operator dashboards reading from an append-only job status store (success / failed / pending) with CSV/Excel export, so operators were not retrying failures that could only re-fail.
  • Engineered data integration monitoring systems using ReactJS as the middleware UI layer between NetSuite ERP, Sailthru, and third-party platforms across retail, finance, and enterprise clients.

NETSUITE · SUITESCRIPT · REACT · SAILTHRU

Selected work

Advance-POS

Multi-tenant restaurant POS deployed across Texas: offline-first Android tablets on a peer-to-peer TLS mesh, so orders and card payments survive an internet outage. Duplicate charges prevented by idempotency keys enforced at both the database and the payment gateway.

+ MORE DETAIL
  • Architected and engineered the full stack (React Native, Node.js, GraphQL, PostgreSQL + Redis) across mobile, web, and backend layers.
  • Offline-first coordination over a peer-to-peer TLS LAN mesh (mDNS discovery, elected owner): order entry, check splitting, card and cash payment, and receipt printing continue through an internet outage.
  • Client-minted UUID idempotency enforced at both the database constraint layer and as the payment gateway transaction id, preventing duplicate charges across offline queue replay and peer-sync conflicts.
  • Optimized schema and query performance (PostgreSQL + Redis), reducing concurrent request latency by 35%.
  • Co-designed a team-wide AI-assisted engineering system (Cursor, Claude Code, agent instructions, TDD workflows) adopted by 3+ engineers.

REACT NATIVE · NODE · GRAPHQL · POSTGRESQL · REDIS

The Beard

Betting analytics and ledger platform for a private trading team. Scheduled odds ingestion computes edge against sharp reference lines, and a fine-tuned GPT-4o-mini parses unstructured Telegram messages into structured picks, in production.

+ MORE DETAIL
  • Scheduled odds ingestion into a MongoDB time series compares each book against a sharp reference to compute price discrepancy, expected value, and recommended bet size.
  • Append-only transaction ledger as source of truth with derived balance caches recomputed on write, keeping balances reconstructible from history and aggregation off the read path.
  • Performance work: denormalized balances, explicit index set, cached connection pool, and odds ingestion offloaded to a queue.

NEXT.JS · NODE · POSTGRESQL · MONGODB · OPENAI FINE-TUNING

Play4Serve

Court booking and management for a pickleball venue in the Philippines. Players hold a slot, pay by GCash or Maya QR and upload proof, and staff verify the payment before the booking confirms. Every staff action is written to a tamper-evident audit log.

+ MORE DETAIL
  • Two surfaces in one Next.js app: a player booking flow and an operator console gated by Supabase Auth against a staff allowlist table with admin and assistant roles, rather than an environment flag.
  • Hash-chained audit log where each row carries a hash of its predecessor, so a removed row breaks the chain. Corrected a linearity defect where rows written in one transaction shared transaction_timestamp() and were ordered by random UUID, splitting the chain into a tree whose dangling leaves could be deleted without detection.
  • Slot holds expire on a pg_cron sweep, so an abandoned checkout releases court time instead of holding it, with separate sweeps for bookings abandoned before payment and for proof that is never verified.
  • The anonymous slot-occupancy API returns availability carrying no PII, so the booking grid renders for signed-out players without exposing who booked.

NEXT.JS · TYPESCRIPT · SUPABASE · POSTGRESQL · PG_CRON

VISIT SITE ↗

River Dinks

Court booking for a two-court venue in Cebu. Players book with no account at all, and double-booking is prevented by the database rather than by application logic.

+ MORE DETAIL
  • Accountless booking: a four-character code drawn from a 32-symbol ambiguity-free alphabet, plus the player's mobile number, are the two credentials for finding and managing a booking. Nobody signs up to reserve a court.
  • Double-booking made structurally impossible by a partial unique index on (court_id, booking_date, hour) across active slots, with a SQL-level test asserting the constraint holds rather than trusting the application path to be the only writer.
  • Booking times stored as a date plus an integer hour in Asia/Manila and never as timestamps, so no UTC conversion or DST rule can shift a reserved slot.
  • Payment by GCash reference submitted by the player and confirmed by the venue, with expiry sweeps releasing holds that were never paid.

NEXT.JS 16 · TYPESCRIPT · SUPABASE · POSTGRESQL · TAILWIND 4

VISIT SITE ↗

HRIS Platform

Cross-platform HR system covering records, attendance, leave, and payroll, with authorization kept server-side behind a Node.js layer and the legacy codebase migrated to TypeScript.

+ MORE DETAIL
  • Interposed a Node.js backend between clients and Firestore, keeping authorization beyond reach of a compromised client, with payroll stored separately from general employee records.
  • TypeScript migration eliminated pervasive "any" from untyped Firestore documents, moving document-shape errors from runtime to compile time.

REACT · REACT NATIVE · NODE · FIRESTORE · TYPESCRIPT

Built on my own time

Ilaw Cove

A complete booking platform for a fictional sixteen-room boutique hotel: availability search, reservation, guest accounts, and an admin surface. Built to work the booking and payment problem end to end without a client's constraints.

GUEST demo.guest@ilawcove.example · ADMIN demo.admin@ilawcove.example · PW Testing123! · RESETS NIGHTLY

NEXT.JS 16 · SUPABASE · POSTGRESQL · RESEND

VISIT SITE ↗

Thessin

Storefront for a fictional apparel studio: category and collection pages, search with filters, and checkout through PayMongo by card, GCash, or Maya. The bag revalidates price and stock against the database every time it opens, rather than trusting what the browser cached.

NEXT.JS 16 · REACT 19 · SUPABASE · PAYMONGO · PLAYWRIGHT

VISIT SITE ↗

Skills

Languages & frameworksTypeScript, JavaScript (ES6+), Java, React, Next.js, Remix, React Native, Node.js, Express.js, GraphQL, Redux Toolkit
Data & storagePostgreSQL, MySQL, MongoDB (incl. time series), Firebase/Firestore, Redis, DynamoDB
Cloud & deliveryAWS (Lambda, S3, Cognito, API Gateway), Docker, Vercel, Jenkins, Bamboo, TurboRepo, Serverless Framework, CI/CD
Enterprise platformsNetSuite SuiteScript (user event & scheduled scripts), Chargebee, Sailthru
AI engineeringOpenAI fine-tuning (GPT-4o-mini), production LLM pipelines, structured extraction, Claude Code, Cursor, Windsurf, agent workflow design, AI-assisted TDD
Practices & testingTDD, Agile Scrum, microservices, REST APIs, WebSockets, Apollo Client, monorepo architecture, performance profiling, Jest, Playwright, React Testing Library, Mocha, BrowserStack, Artillery

Education

Bachelor of Science in Information Technology, Cebu Institute of Technology – University

2014 - 2018

Contact

Interested in working together?

I'm open to new roles and projects, especially where the system has to be right the first time. The fastest way to reach me is email.

Book a callGitHubLinkedInInstagramFacebook