At 10xScale.ai, I built CareerPilot’s Typesense retrieval system to match candidates across 100K+ listings, with a custom resume-to-job-description ranking algorithm. I also architected a Neo4j/Memgraph candidate-ranking graph that cut recruiter screening time by about 70%.
For the Industry Pilot, I built persistent memory, summarization, and RAG retrieval to sustain coherent multi-hour AI teaching sessions, and led the Teaching and Video agents in a four-agent orchestration. I built ResumePilot’s conversational resume builder, which raised in-app job-description alignment scores by 40%+.
I started at 10xScale.ai as a Machine Learning Intern, where I set up Typesense hybrid search and an embedding indexing pipeline, and developed structured LLM assessments for candidate screening. I also built Energetic, a full-stack AI fitness coach, and a prompt-evaluation harness for automated test generation and scoring.
