At Target Corporation, I work across BPX Joy Orchestrator, BPX Search Platform, and BPX Orchestrator Jobs, building production backend and AI systems in Java and Python.
For BPX Joy Orchestrator, I architected an agentic memory subsystem for a Google ADK multi-agent orchestrator, enabling user-scoped context to persist across sessions and agents. I also built a scheduled MongoDB-to-Neo4j ingestion pipeline and an LLM-driven memory promotion engine.
On BPX Search Platform, I led backend architecture for document- and domain-level search and designed an Elasticsearch relevance algorithm. I engineered an AI-powered Summary feature that cut manual document review by ~40%, then re-architected it as an SSE streaming API that reduced perceived latency from ~6 seconds to sub-second.
I also bootstrapped BPX Orchestrator Jobs and contributed to onboarding and chatbot systems. In my SkyFeed project, I led development of a cloud-native video streaming platform with an AWS-based video transcoding pipeline.
