Developer PlatformAccess Management GroupPlatform Access and Identities for Developers (PLAID) teamdevelopers.facebook.comAccomplishments - As a technical leader, launched the Access Levels Feature to over 200 million Facebook apps, providing granular control over access to Meta's data- As a technical leader, designed and implemented two crucial privacy safeguards for over 200 million Facebook apps, ensuring compliance, data security, and support for Meta's privacy commitments through server-side automation and analytics. - As an intern manager, mentored two Facebook interns, contributing to their return offers, through project scoping, pair programming, and code reviews.- Developed a system to maintain consistent state information for Facebook apps, achieving parity between the server layer and data warehouse and reducing instances of data inaccessibility by approximately 95%.- Collaborated with product managers and designers to enhance the Facebook app creation, review, and data use checkup experiences, utilizing new UI components and design languages.

Software Engineer, Egencia Enterprise Business Intelligence Team Accomplishments:- Decreased deployment cycle time for data apps by 80% via a CI/CD framework using stateful containers & DB snapshots - Reduced LOC for error handling & retrying on API calls by 75% via an API consumption framework; led to adoption by 3 teams- Reduced ops time spent on GDPR requests by 75% by developing a GDPR obfuscation service for S3 and Redshift dataSoftware engineer, Egencia Enterprise Analytics Team Accomplishments:- Reduced runtime of CRM data pipelines by 4x via cloud migration from SQL server/PowerShell to AWS/Python- Reduced data silos by 50% by automating cross environment ETLs between Redshift clusters via a python framework- Enabled intelligent table schema versioning & type casting by modernizing the data ingestion layer to use AWS Glue

Product Discovery & Evaluation TeamTechnologies: AWS, ELK stack, Serverless, Kinesis, Docker, Kubernetes, Python, SQL, GoAccomplishments:- Reduced time-to-insight by 10x by building real-time ETL pipelines with Spark SQL to power KPI dashboards - Automated pipeline deployment & testing with 99.99%+ uptime by developing Ansible playbooks & Python libraries- Surfaced and fixed hidden bugs in production by setting up monitoring dashboards with Cloudwatch,Datadog, and ELK stack - Reduced LOC by 75% and added multi-cloud compatibility for deployment scripts of data processing lambdas using ServerlessContract via AIM Consulting

Technologies: AWS, Azure, Google Cloud, JavaScript, RapidMiner, Python, Java, Agile Development Accomplishments: - Created fashion discovery engine with AWS Sagemaker, SQS, and Lambda based on macro fashion trends and user lifestyle - Integrated Clarifai & Microsoft computer vision APIs with Lambda to tag apparel images with labels generated from WordNet- Created synonym search and auto-complete UI features with Elasticsearch & reduced query latency by 30% with ElastiCache- Built frontend & onboarding experience with React Native for Unity app that provides real-time 3D motion tracking

Technologies: C++, Python, Caffe, SQL, OpenCV Accomplishments: - Developed computer vision algorithms to detect buildings in satellite images; achieved 86%+ AUC - Automated data quality control using machine learning to filter out images with few buildings- Open-sourced a dataset of satellite images, shapes and heights of 40K+ buildings in US cities; first dataset of its kind - Implemented neural nets from research papers; improved accuracy by 10% using Bayesian optimization