Work History
Amazon — Smart Vehicles
Software Development Engineer Intern
- —Compress and adapt multimodal foundation models for deployment directly on vehicle hardware.
- —Build the inference stack powering in-vehicle voice assistants, from speech recognition through response generation.
Amazon — Leo (formerly Project Kuiper)
Applied AI Engineer (contract)
- —Forecast space policy shifts across government agencies using an ensemble of time-series foundation models.
- —Structure large unstructured document corpora into clean, deduplicated topic clusters for downstream modeling.
- —Improve signal quality in noisy policy data through temporal smoothing and filtering techniques.
Mercor
Software Engineer
- —Build preference datasets for RLHF alignment of multimodal foundation models on complex tasks.
- —Establish human performance baselines for autonomous agents operating across technical environments.
- —Construct supervised fine-tuning datasets for advanced programming and agentic terminal workflows.
General Dynamics Information Technology
Software Engineer Intern
- —Deploy a computer vision module for real-time vehicle detection within a multimodal ML system at the edge.
- —Harden model robustness against distribution shift through adversarial training techniques.
- —Optimize the inference pipeline for low-latency operation on streaming data using quantization and concurrency.
General Dynamics Information Technology
Engineering Project Lead (contract)
- —Lead a cross-functional engineering team building an agentic RAG pipeline over a large-scale document lake on AWS.
- —Architect a hybrid graph and vector database design to dramatically improve data discoverability.
- —Enhance retrieval quality through custom semantic chunking and entity-aware metadata enrichment.
Booz Allen Hamilton
Engineering Project Manager (contract)
- —Build full-stack tooling to automate medical policy extraction with real-time editing and annotation capabilities.
- —Engineer backend data pipelines and web scraping systems for large-scale document aggregation.
- —Integrate NLP models for compliance automation across summarization, extraction, and sentiment tasks.
Internal Revenue Service
Software Engineer Intern
- —Maintain ETL workflows underpinning enterprise fraud detection and tax compliance systems.
- —Modernize CI/CD pipelines and version control integrations as part of infrastructure uplift.
- —Migrate legacy systems to current build automation frameworks to accelerate deployment cycles.