πΌππππππ π»πππππππ πΏπππππππ
β’Built website [ReactJS, MaterialUI] (Machine Learning Portal) for streamlining deployment of ML models in production environment
β’Built backend webservice (Python, Flask, NoSQL) to communicate with AWS infrastructure
β’Created custom library (Python, Flask, TensorFlow) to provide developers with a toolkit for working with ML models
β’Built AWS infrastructure
β’Deploy and auto-scale microservices w/ Kubernetes
β’Key-Value store for ML model Features w/ DynamoDB
β’Train and inference w/ SageMaker
β’API to microservices w/ API Gateway
β’Asynchronous data pipelines w/ Kinesis and Firehose
π³πππ π΄ππππππππππ
β’Built website [VueJS] (Data Portal) and backend [NodeJS, ExpressJS, TypeScript] to connect internal users to data warehouse in Snowflake
β’Built custom SQL IDE, with IntelliSense and batching, to facilitate exploration and querying of internal data.
β’Saved users' queries and results on cloud
β’Built discoverable, shared query libraries
β’Added custom links for sharing queries and results
β’Built CRON-job scheduler to run queries automatically at times set by users.
β’Created data pipeline to fetch / send trip data to PWC through a Secure File Transfer Protocol on AWS
β’Auto-log user events for analyzing usage
πΏππππππ / πΈπππππππ πππππππ
β’Designed and built new architecture [Ruby on Rails, ReactJS, MySQL] for internal, role-based permissions with ad-hoc permission requests.
β’Built API service [Ruby on Rails, RESTful API] to connect 3rd-party service (Forethought.ai) to various internal services and DB operations
β’Built service [Ruby on Rails] to pass insurance form data from an external webhook (TypeForm) to a 3rd-party service (OrigamiRisk). Before passing, data is parsed, cleaned, and used as reference to fetching / sending additional data from DB.
β’Documentation (API / Services / Coding Best Practices / Onboarding Guidelines)
