• Created 22 aggregated insights based on customer data from 9 different regions to send weekly efficacy reports to internal stakeholders to streamline improvements to AWS services. Created design plan and communicated the ambiguities with principal engineers, managers, and key stakeholders to address stakeholder requirements.

• Created 2 stack infrastructures in TypeScript for configuring AWS resources, i.e., OpenSearch clusters, DynamoDB (DDB) Global Table, SES, Lambdas, IAM (Roles, Policies, Cross-Account Delegation), and EventBridge.

• Configured (3) lambdas/packages in Python for generating aggregated insights from existing OpenSearch clusters in each region and delegating data into a global table for cross-region communication (triggered by cron job), pushing data from DDB stream into a new OpenSearch cluster, and leveraging Elasticsearch aggregation capabilities to generate visual reports using matplotlib to send to stakeholders via SES (triggered by cron job).

• Scaled code to handle ~60,000 customer records, verified code integrity with unit tests (Pytest), and pushed ~6000 lines of code (including tests) into production environment.

• Migrated full text searching from MySQL to Elasticsearch (ES) for 2 data tables to improve performance for keyword and analytical searching. Included +1200 lines of test code to cover UI automation testing and unit testing around ES indexing using RSpec and Selenium.

• Added support for ES bulk reindexing in Rails on transactional processes to optimize performance on transactional updates (upwards of 400%) by reducing multiple calls to reindex data in the ES documents to a single call.

• Improved procurement workflow by enabling support for tagging objects to allow customers to track and filter documents. Led the technical design and communicated with PM and architecture team to discuss performance concerns.

• Added support for tags to the UI, data tables (DTs), API, and CSV loader in Rails. Wrote a migration script to enable ES support for tags on the DTs and covered UI automation and unit test cases (+500 lines).

• Fixed (7) bugs in the procurement backlog to address customer issues found in production.

• Built watchdog service in React to identify and resolve background job

failures in purchase order creation flow.

• Consolidated SQL queries to fetch data into a service wired up to the UI to display diagnostics; whitelisted task in Ruby for support team to resolve failures in the foreground to prevent system backlog and improve P2P flow.

• Built 2 analytics pages in React to provide insights on supplier items (1) and on warehouses by displaying KPIs and comparing instance data to community data (2).

• Added new controllers, models, permissions, services, and data tables (Rails) with low-level caching to each page to optimize performance (tested on customer clones). Wired up a service to an internal API endpoint that ran queries to compute KPIs on community data from S3 to compare against instance data.

• Integrated a 3rd party API into a React app in order to refactor an existing community widget implemented in Ruby.

• Developed a Naive Bayes classifier using Apache Spark's machine learning libraries to predict billing codes on invoices.

• Extended functionality of API services for importing requisitions to inbound/outbound endpoints in Ruby. Tested requests with RSpec and Postman.