At Atlassian, I contribute to enterprise AI search and knowledge-assistant capabilities for Jira and Confluence. I implemented permission-aware hybrid retrieval with citations and authorization checks, improving top-result relevance by about 12% in offline evaluations.
At Practice AI, I led full-stack delivery of legal-medical document intelligence workflows that reduced first-pass case-preparation effort by more than 50%. I also developed document-grounded drafting workflows that cut typical drafting time from 3–5 hours to 30–45 minutes.
At Optum Global Solutions, I developed claims-review products and anomaly models, improving inaccurate-claim identification by 18%. I also implemented clinical NLP services that achieved about 85% precision for common entities.
Earlier, as a Junior Software Engineer at ByteDance, I supported internal operations and content-management tools, including automation scripts that reduced repetitive reporting and data-cleanup work by about 8 hours per week.
