Francen Timothy N. Castro
Profile
I’m a full-stack developer building practical, high-quality digital products that solve real problems.
I work across the stack, using React, Next.js, or Vue.js on the frontend to create clean, responsive interfaces, and
Node.js, PHP (Laravel), or Python (FastAPI, Flask) on the backend to build secure, scalable systems.
I’ve developed a wide range of projects, including marketplaces, dashboards, eCommerce platforms, and realtime applications, always focusing on performance, reliability, and a smooth user experience.
I’m comfortable working with MySQL, PostgreSQL, and MongoDB, and I use Docker, AWS, and CI/CD pipelines to
make deployment fast and dependable.
Professional Experience
04/2022 – 08/2025
Houston, Texas, US
Nexxt Ideas
Software Developer
Delivered responsive, pixel-perfect user interfaces from Figma designs using React
and Next.js, improving cross-device consistency and performance by 40%.
Collaborated with product, backend, and design teams to build a scalable operations
dashboard supporting 5,000+ active users, reducing task processing time by 25%.
Architected a modular frontend system with Redux and Redux Toolkit, reducing
redundant code by 30% and improving long-term maintainability.
Integrated 10+ RESTful APIs to power real-time analytics, dashboards, transactions,
and user activity tracking, improving data delivery speed by 45%.
Optimized frontend performance using lazy loading, memoization, and code
splitting, cutting average page load times by 50% and increasing user retention by
18%.
09/2018 – 03/2022
Austin, Texas, US
Directio
Software Developer
Improved application scalability and performance by 15% through Next.js upgrades
and Redux Toolkit migration.
Built and launched a blockchain-based crowdfunding platform using React,
TypeScript, and Web3.js, onboarding 1,000+ investors and processing over $500K in
transactions.
Implemented automated testing using Jest and Cypress, reducing post-release
defects by 50%.
Developed and maintained a reusable React component library (200+ components),
accelerating feature development by 25%.
Deployed containerized applications using Docker, reducing release cycles by 35%
and improving uptime.
Integrated Django CMS to streamline content management and reduce build times
by 30%.
Automated CI/CD pipelines, increasing deployment frequency by 20% and reducing
manual release effort by 30%.
01/2016 – 12/2017
Chicago, IL 60604, USA
Ironbeam
Software Developer
Led the migration from a monolithic architecture to microservices, improving
system reliability by 45% and reducing infrastructure costs by 30%.
Designed custom caching and load-balancing solutions using Redis and Nginx,
reducing API latency from 1.8s to under 600ms.
Coordinated agile sprints across frontend and backend teams, achieving 98% sprint
completion rates.
Enhanced CI/CD pipelines with Jenkins and GitLab, reducing rollback incidents by
40%.
Improved code quality and onboarding efficiency through documentation and static
analysis, reducing ramp-up time for new engineers by 35%.
Projects
Sourcefy Marketplace
This large B2B product marketplace connects buyer and seller companies globally.
The platform includes three user roles: admin, buyers, and sellers.
As an admin, users can manage dashboards featuring transaction data, GMV, revenue, client and supplier details,
product information, and financial graphs.
For buyers, the system enables project postings and proposal management, while suppliers can review projects
and send proposals.
I designed several APIs for user management, transaction processing (integrated with Stripe), messaging via
WebSocket, and dispute management.
The technical stack comprised React.js on the frontend, Nest.js on the backend, PostgreSQL for data storage, and
Docker with GitHub Actions for CI/CD.
RAG Chatbot
I developed a Retrieval-Augmented Generation (RAG) pipeline for legal research, leveraging LangChain, GPT-4,
OpenAI embeddings, and Pinecone vector databases to deliver contextually relevant legal information. The
system integrates Google Search to supplement retrieved data, enhancing both coverage and accuracy. I
designed and managed two dedicated Pinecone indexes—one for laws and regulations and another for legal
cases—enabling efficient and precise retrieval of relevant content. Using LangChain’s agent functionality, the
pipeline intelligently determines the most appropriate tool to handle each query, allowing the system to
dynamically retrieve and synthesize legal information, ultimately streamlining research and decision-making for
users.
Education
08/2010 – 03/2014
Williams College
Languages
English
Skills
React.js, Vue.js, Angular.js, Next.js, Javascript, PHP, Python, Node.js, React Native, Flutter, Nest.js, Mysql,
RabbitMQ, Postgresql, CI/CD, Docker, Dynamodb, AWS Lambda, Chatbot, LLM
