AVIV ROTH

Architect, implement, and deploy REST APIs to support Machine Learning team's model training and inference analysis on water sound data. Implement and maintain Terraform of AWS infrastructure of water flow monitoring AI system. Write schema and data migrations. Interview engineering candidates. Mentor junior engineers. Audit and improve software life cycle processes. Evangelize software testing. Write GitHub Actions to run unit tests and deployment Terraform in CI/CD.

• Migrated production database to more robust and secure RDS instance, in a private subnet, behind an SSH Tunnel, separate from development database.
• Built network infrastructure on AWS with private VPC and restricted SSH tunneling to database and resources, securing private resources from the internet.
• Integrated entire existing AWS infrastructure with Terraform as Infrastructure as Code (IaC), allowing infrastructure changes to have their histories in GitHub and to be easily rolled back.
• Created SQLAlchemy-like database migration platform, for AWS Lambda codebase using ORM-free raw SQL. Database schema changes now have a change history in GitHub and can also be easily rolled back.
• Wrote PyTest fixture to mock production databases inside unit tests. Code, including raw SQL code, is now better tested and more robust. Production databases are no longer vulnerable to SQL tests gone awry.
• Implemented backend and REST API for "Control Center", application used by Machine Learning team to manage AI model training, inference, and insights from water sound data. Enabled better UX for Machine Learning team and protected the production database from direct SQL queries.
• Optimized Machine Learning team's training queries to run without timing out.
• Developed training parameter validation system for Control Center, allowing rules to be written in YAML and to be modified quickly and easily.

Architected and implemented features and tests for REST API backend for shopping cart product warranties, using Python/Flask/Marshmallow/Alembic/PostgreSQL. Wrote REST API documentation. Reviewed and approved pull requests. Refactored and tested legacy code and other tech debt. Assisted customer support and finance.

• Implemented backend for Real Time Claims Chatbot, saving hours of customer support work a day.
• Refactored and expanded claims handling API to improve UX for both customers and customer support.
• Created greenfield backend for self-insuring initiative, allowing warranties to be sold at higher margins and lower prices.
• Increased unit test code coverage by 13-23% on various modules of the product offering.

Implemented features and tests for Django REST Framework API for student off-campus rental platform.

• Built image compression system, speeding up page loads by up to 75%.
• Integrated Salesforce with core platform, allowing easy tracking of rental applications and renter support.
• Implemented and installed security safeguards, including secrets management.
• Delivered roommate invitation module, balancing legacy code reuse with new development.

Developed ETL pipelines for integrations in Python/Django/PostgreSQL, using Test Driven Development. Pair programmed with junior and senior developers. Worked with Product Management to manage project priorities.

• Successfully delivered key integrations on schedule for three major retailer clients.

Developed, tested, and maintained SaaS platforms for marketing intelligence, primarily for analysis of online advertising data (e.g. Google Analytics, DoubleClick Campaign Manager) using Python/Flask, JavaScript/NodeJS. Maintained and developed new features for legacy web sites and custom marketing applications in C#.NET. Communicated with clients to understand software requirements and relay progress.

• Implemented initial version of data ingestion module for flagship SaaS marketing platform, including interfaces to Google APIs (DCM Search, Display), and OAuth support.
• Deployed and configured all initial subnets and servers on AWS for primary SaaS application, including web servers, RDBMS, Load Balancers, SSL, and middleware applications, on both Linux and Windows.