Spearheaded the development of a program designed to enhance activation rates and attract top-tier AI coding trainers, ultimately refining our clients' LLM coding skills, reaching an increased rate of 23%

Constructed graphs for seamless data tracking by using Snowflake to query our schema in Redash, adopting several different databases commonly used throughout the company.

Leveraged Replit, a simple self-hosted repository platform, to optimize recruitment cost and the skill evaluation processes, successfully eliminating the need for paid services like HackerRank while achieving the same high-quality results.

Reviewed data labelled code submitted by contributors for any mistakes or hallucinations made by the AI model to ensure the best possible data quality for client

Implemented technical solutions such as using Google Apps Script to automate the processing of incoming Google Form data, streamlining the workflow by determining whether submitted issues required automatic escalation to the operations team.

Guided users though the platform, either addressing UX confusions, or escalating abnormalities to the operations team