At ALKAMI TECHNOLOGY, I developed an LLM-powered developer platform that automates banking SDK component creation, verification, review, and approval. First-pass validation improved from 72% to 85% in internal beta evaluation.
I also engineered LangGraph-based coding agents and conversational banking agents. Across 300 multi-turn scenarios, conversational agent task completion improved from 84% to 91%; orchestration and recovery optimizations reduced cost per successful task by 22%.
For enterprise knowledge work, I designed a RAG Knowledge Assistant using Python, FastAPI, LangChain, AWS Bedrock, PostgreSQL/pgvector, and React. Hybrid retrieval and citation-aware generation improved grounding acceptance from 78% to 90% across reviewer-scored questions.
Before my AI engineering work, I built data and integration services at ALKAMI TECHNOLOGY and enterprise workflow and document-processing systems at IBM. My work included banking APIs and event pipelines, as well as IBM Datacap OCR and FileNet integrations.

