Christian T. Dataro

Delivered RoundUp end-to-end - a regulated real-money digital lottery platform spanning five game types (round-up, instant-win, mega-jackpot, mystery-box, and subscription draws) across user-web, admin, kiosk, and mobile channels.

• Architected a 29-module NestJS 11 backend (32 TypeORM/PostgreSQL entities) split into five purpose-built AWS Lambdas - API, async processing, notifications, webhooks, migrations - for independent scaling and deploys.
• Engineered an event-driven draw lifecycle (EventBridge + SQS) that keeps exactly one draw open at a time and auto-allocates prizes on close - eliminating manual draw operations.
• Authored lpfy, a zero-dependency Lazy Partial Fisher-Yates library for CSPRNG-backed winner selection with rejection sampling to eliminate modulo bias, in O(m) memory across billion-scale ranges.
• Shipped three React 19 clients (player, admin, kiosk-operator) on Vite/Tailwind 4/Zustand/TanStack Query, plus a React Native app (RN 0.79) with biometric auth, QR scanning, and push.
• Hardened the platform for regulated real-money play - AML/KYC, MFA (SMS/Email OTP + TOTP), play/deposit/session limits, and six-role RBAC via AWS Cognito.
• Integrated Stripe payments/webhooks with the internal wallet, plus Firebase push, SendGrid email, and QR kiosk cash-in.
• Provisioned multi-region AWS (eu-west-1/2, us-east-1) from 22 reusable Terraform modules across 58 Terragrunt stacks; guarded by SonarQube static analysis and 110+ Playwright end-to-end test cases with Mailosaur OTP verification.

Projects: RoundUp - Digital Lottery & Gaming Platform

Stack: NestJS 11, TypeScript, TypeORM, PostgreSQL (Aurora), Stripe, AWS (Lambda, Aurora RDS, S3, CloudFront, Cognito, SQS, KMS, Secrets Manager, SES, EventBridge), Terraform, Terragrunt, React 19, Vite, Tailwind CSS 4, Zustand, TanStack Query, Radix UI, React Native 0.79, Firebase, SendGrid, OpenAPI (Orval-generated typed clients), Playwright, Mailosaur, accessibility testing (axe-core), Lighthouse CI, Sentry, Jest, lpfy

Lead maintainer across GlobalNet's iGaming platforms - a live micro-betting sportsbook, a multi-vendor provider-integration layer, and a casino operator suite - modernizing legacy code and setting engineering standards.

• Built the live micro-betting client (Next.js 16 / React 19) across eight feature modules, streaming per-game state over a NATS JetStream WebSocket on a sport.zone.league.gameId subject into a Zustand store.
• Contributed five Spring Boot (Java 21) services (games, wagering, wallet, backoffice, feed transceiver) and an nVenue SSE-to-NATS transceiver with per-sport translation strategies for basketball, baseball and American football, modeling the full bet lifecycle (open, paid, resulted, settled, failed, voided) over PostgreSQL/Liquibase.
• Documented the platform as code - a C4/Structurizr model, 18 PlantUML sequence diagrams across six services, an architecture decision record (ADR) on active-market snapshots, and root-cause analyses on React 19, CI/CD, API mapping, and test strategy.
• Standardized shared UI via a Webpack Module Federation package that ships React, Next.js and Vue adapters from one codebase, plus a shared sports library (reconnecting NATS client + TanStack Query hooks).
• Unified 12 third-party gaming providers (Playtech, Forge, BGaming, EvoPlay, RTG, JDB, Alea, Wager2Go and more) behind one seamless-wallet contract - nine inbound proxies exposing 46 endpoints plus three outbound client libraries.
• Implemented per-provider request signing (HMAC, MD5/SHA-512 digest, or AES-encrypted action payloads), and hardened operator auth with a license-key validation plugin over cross-module reporting and audit.
• Audited the 12-service legacy integration estate and documented systemic risk per provider - committed secrets with no vault pathway, weak MD5 signing, outbound HTTP inside transaction boundaries, error-masking fallbacks on credit and rollback, and end-of-life dependencies.
• Modernized the casino operator estate - migrating the Vue 2 Options API admin to Vue 3 Composition API (Vuex 4, PrimeVue 4, AG Grid 32) across 148 components and 43 routes, and rebuilding legacy Phaser slot games into React 19 + TypeScript on a Figma-designed portrait-mode UX.
• Guaranteed quality with 592 JUnit tests (29 of them Testcontainers-backed against real PostgreSQL and NATS), 89 Cypress end-to-end cases, and 155 Vitest unit tests; stood up Docker CI/CD for staging and production.

Projects: SportsBetting - Live Micro-Betting Platform; SportsBetting BackOffice + API Integrations; Casino Platform - Game Center Admin & Slot Games

Stack: Next.js 16, React 19, Vue 3, Phaser, TypeScript, TanStack (Start/Router/Query), Spring Boot 4 (Java 21), Zustand, Vuex 4, NATS JetStream, Webpack Module Federation, PrimeVue 4, AG Grid 32, Chart.js, Fastify, PostgreSQL, MySQL/MariaDB, Liquibase, Cypress, Playwright, Vitest, JUnit + Testcontainers, Structurizr (C4), PlantUML, Docker, legacy PHP

Refactored a veteran career platform - job search, applications, e-portfolio and AI resume building, skills assessments, a video interview simulator, federal WIOA compliance reporting, and caseworker case management - serving 15 US-state workforce agencies, from a legacy codebase into 3 configurable Turbo monorepos (6 apps: 3 NestJS APIs + 3 Next.js frontends).

• Refactored a legacy codebase into 3 configurable Turbo monorepos (job-seeker, employer/provider, case-management) serving 15 US-state workforce agencies, cutting build and maintenance overhead ~30%.
• Designed a NestJS 9/10 + MongoDB backend spanning 130+ modules, 150+ Mongoose schemas and 154 Swagger-documented controllers exposing roughly 1,170 REST endpoints across the veteran career domain.
• Shipped federal WIOA compliance reporting - PIRL participant-record generation, soft-eligibility screening, a review-board workflow, and SFTP delivery of generated submission files - alongside 14 scheduled jobs driving notifications and eligibility.
• Built the AI feature set on a provider-swappable OpenAI layer (GPT-4) powering resume generation and job matching, plus a multi-step video interview simulator with in-browser MediaRecorder capture, per-step S3 upload and FFmpeg-generated thumbnails.
• Secured multi-tenant access with a CASL ability matrix of 14 actions across 18 subjects over database-driven roles, AWS Cognito user pools verified by JWKS/JWK-to-PEM middleware, and field-level encryption at rest for SSN and jobseeker PII.
• Built a no-code report and dashboard builder (SQL-to-MongoDB querying, drag-and-drop layout, D3) on dynamic Form.io forms with server-side DOCX, XLSX and PDF generation.
• Owned delivery on AWS EKS - per-service Terraform stacks with horizontal pod autoscaling, WAFv2 and Secrets Manager, driven by 14 GitHub Actions workflows across branch-gated dev, QA, UAT and production lanes with ECR image builds and SonarQube static analysis.
• Guaranteed quality with 225 Jest spec files across the six apps, API suites running against in-memory MongoDB, and Prometheus metrics plus Winston-to-CloudWatch logging in production.

Projects: Veteran Career Platform

Stack: Next.js 13, React 18, TanStack Query, Redux Toolkit, MUI (X Data Grid), TypeScript, NestJS 9/10, MongoDB, Mongoose, CASL, OpenAI (GPT-4), AWS (Cognito, EKS, SQS, SNS, Secrets Manager, CloudWatch, WAFv2, S3, ECR), Terraform, Kubernetes, Helm, Redis, Prometheus, Winston, Form.io, FFmpeg, docx, ExcelJS, Puppeteer, @react-pdf/renderer, SFTP, Swagger, Turbo, Jest, mongodb-memory-server, GitHub Actions, SonarQube, Docker

Built Admin Console (CMS for users, super admin, marketing, tester, customer) and Tenant Portal (transactions, auto-pay, card payments, warehouse rentals) for a US warehouse-rental business.

• Built an Admin Console (CMS for users, super admin, marketing, tester, customer) and a Tenant Portal (transactions, auto-pay, card payments, warehouse rentals).
• Integrated a Pay-and-Bill payment gateway, CRM customer management, and IAM authorization; implemented the Supabase data layer.
• Coordinated requirements with US Business Analysts, Chinese Managers, and PMs; provided off-hours on-call production support.

Projects: Cubework US - Admin Console; Cubework US - Tenant Portal

Stack: React, TanStack (Start/Router/Query), TypeScript, ShadCN, BitBucket, jwt, ESLint, pnpm, Vite, Vitest, Zod, Jenkins, Supabase

Built and maintained 6 React single-page apps across two product lines (player, agent and admin/operator portals) for real-time betting and agent-network gaming - low-latency live video, SignalR realtime updates, and multi-role operator control.

• Delivered real-time odds, bets, wallet and payout updates over a dual-channel Microsoft SignalR design - a public broadcast feed plus a per-user private feed with subscription registration and automatic reconnect - across six React SPAs and 334 API client modules, on Redux + Redux Thunk state.
• Reduced UI latency ~20% by adding request caching, refetching, and optimistic updates to high-traffic API calls, and pushed session revocation over the realtime channel.
• Enforced multi-role operator access across 7 roles (super admin, overall admin, technical admin, loader, declarer, auditor, CSR) spanning 105 role-scoped routes, plus player and guarantor types, with an activity audit-log surface.
• Integrated nanocosmos NanoPlayer low-latency live streaming into the player betting experience and built the operator control plane for it - per-arena stream provisioning, game-room management, winner declaration, and payout/commission settlement.
• Built a multi-level agent hierarchy with downline reporting and commission settlement, plus a GSAP-animated chroma-key broadcast overlay for live event production.

Projects: IN BETWEEN - Agent System & Pinoy Card Game; UNODOS - World Sabong Cup & API for Korea; UNODOS - Horse Racing Game

Stack: React 18, JavaScript, Redux, Redux Thunk, React Router 6, Microsoft SignalR, CoreUI Pro 4, Sass, GSAP, nanocosmos NanoPlayer (Bintu), FullCalendar, Axios, BrowserStack - backend: Django, MongoDB

Deployed ML models and data products on distributed, 60+ member open-collaboration teams across four projects.

• Improved forecasting accuracy by preprocessing data with Pandas/NumPy and hypertuning ARIMA time-series models (statsmodels), served via FastAPI, and built recommendation engines plus predictive and sentiment models across four ML chapters.
• Shipped ML for four domains - hardware-store inventory demand, course evaluation and topic recommendation, student-debt risk analysis, and student career recommendation.
• Collaborated across 60+ member distributed teams, aligning on functional requirements, scope and timelines.

Projects: Omdena - AI / ML Chapters

Stack: TensorFlow, PyTorch, scikit-learn, Keras, Pandas, NumPy, Polars, NLTK, spaCy, Transformers, OpenCV, SciPy, Statsmodels, FastAPI

Built MDCC Intranet, an internal employee e-learning and training portal with an administrator CMS.

• Delivered course and content delivery for employees alongside a built-in CMS for administrators to author, publish, and manage learning materials.

Projects: MDCC Intranet - Employee E-Learning Platform