Pedro Shin

Background

About Pedro Shin

Brazillian and Italian citizenship, C2 ESOL English https://pedroshin.dev Software engineer with 3+ years professional experience in fullstack (front + backend) development, as well as data science, machine learning and artificial intelligence. • +- Extensive experience with React, Next.js, Typescript, SingleSPA (microfrontends architecture), tRPC and GraphQL in backend-for-frontend (BFF) patterns, Turborepo- Created bespoke in-browser running machine learning algorithms in Python- Custom chatbots communicating with OpenAI's GPT-4 (text), Dall-E2 (image), and ADA-002 (embeddings) APIs- Websocket chatrooms for live message exchange- 3D web page in Three.js displaying cloud of hyperdimensional word vector embeddings with t-SNE dimensionality reduction algorithm- For styling and css, I have worked with a myriad of technologies, such as: Storybook, MUI, tailwind, styled-components, sass, and many others. • +- TypeScript and Python services running on Dockerized cloud servers, in both corporate (production) and personal environments, as well as a multitude of architectures and paradigms- Development of several production level APIs integrating MongoDB in NestJS event-driven microservices architecture- Creation and maintenance of PostgreSQL databases in AWS RDS and Supabase for vector embedding studies, and AWS EBS Dockerized servers running Flask (Python) Websocket APIs communicating with those databases. • - Mathematical understanding of AI/ML algorithms and statistical models- Handmade AI/ML algorithms, such as Genetic Algorithm and K-Means, from scratch using only core libraries (pandas, numpy, matplotlib)- Experience with Python tools for AI/ML, such as scikit-learn and TensorFlow. • /+- CircleCI integration with AWS - Github Actions + Github Pages - AWS EC2, EBS, S3, Cognito, Textract, Rekognition • - React Native and Expo Go personal application published to App Store and Google Play though not publicly available. • - Participated in the core team of engineers and designers that structured my company's internal Design System. This system includes layouts, tokens, components, etc, and was published publicly using Storybook and made available for internal use as an npm package.