Intertru

As a Backend/AI Engineer at Intertru (Newport Beach, CA), I optimized system performance and context accuracy across our AI-generated interview platform. I engineered an in-session retrieval architecture (RAG) that dynamically injects document contexts, improving analysis-point accuracy by 2.3x, and hardened the AI integration against prompt injection using input sanitization and structured JSON enforcement layers. To optimize infrastructure, I refactored legacy background-threaded workloads into asynchronous AWS Lambda calls via Boto3 to eliminate idle EC2 runtime charges and cut operational costs by 20%. Additionally, I increased team developer velocity by building an internal, in-platform asynchronous feedback and archive feature that eliminated 2+ hours of synchronous sprint review overhead per cycle.