Career Summary
4+ years of full-stack development experience specializing in React ecosystems, accessible UI component libraries with design token systems, microservices architecture, RESTful APIs, and performance optimization using caching and content delivery.
Skilled in Agile end-to-end product development, from design and testing through deployment and maintenance.
1+ year experience in AI-integrated application development, deploying large language models(LLMs) on cloud platforms with customized prompt strategies and integrated external tools.
Proficient in AI-assisted development with 2 years experience using AI coding agents (Copilot, Cursor, Windsurf, Claude Code, Gemini CLI, Codex) to accelerate development and improve quality.
Company
Mosaic Manufacturing
Title
Full Stack Engineer
Stack
React, TypeScript, Redux, Node.js, Express, Jest, Playwright, DynamoDB, AWS, Docker
Location
Toronto, Canada
Duration
Nov 2021 - Nov 2025
Overview
As a cornerstone of Team Canvas, my contributions over 4 years include elevating our codebase with over 95% test coverage, completing two comprehensive UI revamps, and executing numerous view redesigns. I led the development of a company-wide UI component library and established containerized testing infrastructure with comprehensive E2E and unit test coverage.
I spearheaded the development of the Canvas Teams feature, released in Summer 2023, now widely adopted by organizations and universities such as World Emblem International, Amazon, BMW, Toronto Metropolitan University, Carnegie Mellon University, Hong Kong University, and McMaster University.
Results and metrics
Drove UI/UX redesign of the company's core web application with cross-functional teams, implementing responsive component architecture with React, Redux, and Styled Components, documented through Storybook
Led the development of an accessible UI component library using Vite and Radix UI, following WCAG 2.1 AA guidelines with ARIA and keyboard navigation, validated through Lighthouse; migrated from styled-components to vanilla-extract for React Server Component compatibility, featuring a custom design tokens system
Developed touch-enabled user interfaces for on-device applications and established Docker-based testing infrastructure with Playwright E2E tests integrated into CI/CD pipeline with complete unit test coverage
Achieved over 95% test coverage through automated Unit and Visual Regression testing using Jest and React Testing Library; integrated Sentry for real-time error monitoring, significantly improving bug detection and resolution times
Architected scalable 3D slicing pipeline with AWS API Gateway (rate limiting, request validation) and serverless backend ( Lambda, SQS, S3) to handle compute-intensive asynchronous processing for concurrent institutional users
Implemented enterprise authentication system with Single Sign-On and role-based access control using AWS Cognito, enabling universities and corporations to integrate with their existing identity providers
AIkie - AI-Integrated Stock Analysis Application
Client
Next.js, React, TypeScript, Tailwind
Server
Prisma ORM, Supabase, PostgreSQL, WebSocket, AWS
AI & Payment & Cache
DeepSeek R1 & V3, Langchain, Stripe, Redis
Authentication
Google & GitHub OAuth
Overview
AIkie is an advanced stock analysis application that integrates AI technology to provide users with deep market insights. Built on the powerful Next.js framework, the platform seamlessly integrates DeepSeek R1 and V3 models, offering real-time stock analysis, market trend predictions, and investment strategy recommendations. By combining various financial data APIs and AI analytical capabilities, AIkie provides investors with a one-stop market intelligence center.
Features & Technologies
Architected full-stack stock analysis platform using Next.js with SSR/ISR/SSG and dynamic metadata for SEO, Supabase OAuth, Stripe webhooks for subscriptions, and type-safe data layer with Prisma ORM and Zod validation
Integrated DeepSeek R1 (chain-of-thought) and V3 models with Langchain orchestration: real-time data fetching from financial APIs, context injection with structured prompts, and domain-specific analysis framework
Built real-time streaming architecture using AWS Lambda@Edge and API Gateway to handle concurrent LLM requests, delivering token-by-token reasoning for enhanced user transparency
Optimized performance with Redis caching layer featuring intelligent TTL policies, achieving sub-100ms response times for repeated queries and reducing API costs
Utilized Langchain to build financial data processing pipelines, extracting and analyzing market trends, company reports, and relevant news, providing users with comprehensive stock insights
Implemented real-time stock data updates and immediate delivery of AI analysis results through WebSocket, ensuring users have access to the latest market dynamics
Managed application state efficiently with Zustand, providing a lightweight and flexible state solution for complex stock filtering, comparison, and tracking features
