Where I've worked

AI Engineer Intern

Worked in the Office of the Chief Data Officer on document classification and metadata discovery across regulatory systems.

• Expanded document-classification coverage 8× by uncovering 9 untracked source subsystems in structural and free-text metadata.
• Validated candidate document types through 50+ clustering, embedding, and LLM-as-judge experiments, using stratified sampling to confirm each held across populations.
• Built a rulebook dependency graph at an internal createathon: 50 rules linked through structural parsing and NLP cross-reference extraction, with a chat interface to trace a rule change's downstream impact.

AI Software Engineer Intern

Built remediation-priority scoring and anomaly detection for security teams.

• Cut mean time to remediate by 20% with a remediation-priority model combining CVSS severity and asset criticality.
• Cut incident response time 30% across 100K+ event streams with an Isolation Forest anomaly detector on AWS GuardDuty/Macie logs.
• Prototyped an agentic system that explains risk scores by querying the security database with LLMs.

Software Engineer

Embedded and tooling work for in-vehicle systems, shipped at real automotive scale.

• Shipped Bluetooth features in Java on embedded Android, deployed in 100K+ vehicles.
• Sped up integration cycles 40% with an in-vehicle network emulator for Bluetooth hardware-in-the-loop testing.
• Raised code coverage 36% (Coverity) and cut development time 20% by introducing LLM-driven unit testing.

Software Engineer Intern

• Cut feature-testing time 50% by building a Bluetooth test application.
• Validated client hardware at 96% compliance with quality benchmarks through sanity testing.

Machine Learning Engineering Intern

• Built a Python backend that migrated graph datasets into Neo4j.
• Automated ingestion with logging, structured schemas, and query optimization.