Taurai – Symfony, Vue.js, Java

An AI-powered social media automation platform that enables teams and creators to generate, review, manage, and publish content across YouTube, Instagram, and TikTok from a single workspace. The platform supports persona-driven content creation, approval workflows, media management, analytics, and secure third-party integrations, helping streamline content operations at scale.

• Architected and developed the full product across frontend, backend, database, and deployment layers;
• Built AI workflows for text, image, and video generation with queue-based processing and moderation/review pipelines;
• Implemented publishing flows for YouTube, Instagram, and TikTok with OAuth-based account connection and token management;
• Designed workspace-scoped multi-tenant architecture for teams managing multiple brands or creator identities;
• Developed profile and persona management to generate and publish content in the voice and identity of assigned creators;
• Built review, archive, and publishing systems for draft approval, regeneration, media history, and auditability;
• Implemented S3-compatible media storage for generated assets and production-ready delivery pipelines;
• Added database-managed app credentials, AI provider settings, static content, and compliance/legal configuration;
• Integrated MCP support to securely expose application capabilities to external AI clients;
• Improved production reliability with structured logging, sanitized error handling, and containerized deployment workflows.

An AI-assisted trading platform for MT5 focused on forex and commodities scalping, primarily targeting XAUUSD and major currency pairs. The system combines multi-timeframe market analysis, Elliott Wave-based pattern detection, regime filtering, confluence scoring, live news monitoring, and AI-driven trade evaluation to generate and manage trading setups with built-in risk controls, execution safeguards, and fallback mechanisms.

Key Accomplishments

• Designed and delivered an end-to-end MT5-to-backend trading workflow for live signal evaluation;
• Built an AI-first execution model where OpenAI generated trade direction, entry, stop-loss, take-profit, invalidation, and expiry parameters;
• Implemented local fallback logic to maintain safe trading operations when AI responses were unavailable, untrusted, or delayed;
• Added top-down multi-timeframe Elliott Wave analysis across H4, H1, M30, M15, and M5;
• Introduced pending limit-order execution and automatic cancellation for invalidated setups before order fill;
• Built runtime controls for risk management, confluence thresholds, AI hold behavior, refresh intervals, and live-trading toggles;
• Reduced AI operating costs through prompt reuse, compressed MT5 payloads, cached AI context, and scheduled request execution;
• Added telemetry, usage tracking, AI latency and token metrics, OpenAI health monitoring, and MT5 log ingestion for debugging and observability;
• Implemented live news integration and removed untrusted or manual signals from production scoring;
• Created MCP and OAuth support to securely expose platform capabilities to external AI agents.

Responsibilities

• Architected the backend signal engine and MT5 execution workflow;
• Designed trade logic, confluence scoring, regime filters, liquidity checks, and risk controls;
• Built AI prompt design, response parsing, caching, timeout handling, and fallback strategy behavior;
• Integrated MT5 market snapshots, broker execution logic, and trade journaling;
• Implemented database models, runtime settings management, telemetry, and admin/debug endpoints;
• Managed local development deployment workflows with Docker and deployment compatibility with Traefik;
• Tuned the system for lower latency, reduced token consumption, and safer live-execution behavior.

A multi-tenant ecommerce and content management platform that enables operators to create branded storefront applications, manage content and products, configure payments, customize themes, and publish customer-facing storefronts. The platform streamlines the entire commerce lifecycle, from application setup and storefront customization to checkout, donations, and public API integrations, helping tenants launch and manage online businesses efficiently.

Accomplishments

• Built a complete go-live workflow from application creation through storefront publishing and payment acceptance;
• Implemented a multi-tenant administration system with app-specific setup, themes, content, products, categories, and payment settings;
• Created reusable public APIs that enabled storefront applications to consume content and commerce data consistently;
• Added payment-session lifecycle support for both purchases and donation flows;
• Built theme-driven storefront rendering with support for imported demo themes and database-backed templates;
• Refactored Docker infrastructure to support both local development and Traefik-based remote deployments;
• Resolved admin dashboard rendering issues by restoring proper compiled asset loading for the Vue-based admin application;
• Improved operational clarity by separating local development configuration from remote deployment settings;
• Established engineering standards, task-tracking practices, schema audit processes, and recovery documentation.

Responsibilities

• Designed and built the core multi-tenant application architecture;
• Developed the admin dashboard for application setup, content management, payment configuration, publishing, and storefront operations;
• Built reusable public Web API endpoints for pages, posts, products, categories, carts, checkout, donations, and user workflows;
• Implemented storefront theme support, including database-backed Twig templates and demo-theme imports;
• Integrated provider-neutral payment flows for purchases and donations;
• Created Docker-based local and remote deployment workflows, including local development environments and Traefik-compatible hosting;
• Managed Doctrine entities, repositories, migrations, and application data models;
• Improved admin UX by replacing raw configuration screens with task-oriented forms and selectors;
• Added security and session metadata handling for sensitive flows such as login, signup, carts, and token issuance;
• Built and maintained frontend admin modules using Vue, TypeScript, and API service layers;
• Debugged and resolved routing, asset loading, admin dashboard rendering, and Docker runtime issues.

A mobile-first dating and meetup platform designed to help users transition from online discovery to real-world connections more efficiently than traditional swipe-based applications. The platform combines profile onboarding, location-based matching, venue-driven meetup scheduling, messaging, subscriptions, notifications, and user verification workflows to support safer and more meaningful interactions.

• Founded, designed, and engineered the product end-to-end, from backend architecture and API design to mobile application development and deployment;
• Built the Symfony backend with authentication, user profiles, match preferences, meetup scheduling, subscriptions, verification, notifications, and chat APIs;
• Developed the Vue 3 and Capacitor mobile application using reusable components, Vuetify UI, Composition API patterns, Pinia state management, and native mobile integrations;
• Implemented multi-provider authentication with Facebook, Apple, Google, email, and SMS login while maintaining secure JWT-based sessions;
• Designed profile onboarding and discovery workflows supporting location-aware matching, interests, preferences, media uploads, and meetup intent;
• Integrated location services, push notifications, WebSockets, media management, and subscription/payment flows across frontend and backend;
• Built safety-focused verification workflows, including configurable verification requirements and extensible selfie/liveness verification capabilities;
• Managed production infrastructure, deployments, server troubleshooting, cache updates, configuration management, and containerized environments;
• Performed forensic debugging using Symfony logs, Xcode logs, simulator screenshots, request payloads, database traces, and frontend build outputs;
• Refactored legacy frontend code toward Composition API and Pinia while maintaining compatibility with existing backend contracts and production data.