Software
Engineer,
AI/ML

Building production ML systems at the intersection of healthcare data, large language models, and applied deep learning. Currently at XponentL (Genpact) and finishing an M.S. in Data Science at UVA.

At a glance
5+
Years building data & ML systems
4.0
GPA, M.S. Data Science, UVA
AWS
Certified Cloud Practitioner & AI Practitioner
Aug '26
Graduation & open to new roles
Python PyTorch PySpark Databricks LLMs AWS Azure MLflow

About me

I'm a Software Engineer focused on AI and machine learning, with a background that spans healthcare data pipelines, applied computer vision, and LLM-powered systems. I care about building things that work in production — not just in notebooks.

Before moving into AI/ML engineering, I spent six years as a Data Analyst in private equity, which gave me a strong foundation in working with messy, high-stakes data and communicating technical findings to non-technical stakeholders.

I'm currently completing an M.S. in Data Science at the University of Virginia (4.0 GPA, graduating August 2026), with coursework focused on deep learning, Bayesian statistics, and NLP. I'm actively exploring research roles and positions at the frontier of applied AI.

Currently
Software Engineer, AI/ML
XponentL Data (Genpact) · Remote
Education
M.S. Data Science · UVA
B.S. Finance · University at Buffalo
Location
Alexandria, Virginia
Open to remote · willing to relocate
Interests
Rock climbing · Running
Chess · Exploring new technology

Selected projects

Deep Learning
ISIC Skin Lesion Classifier
Multi-class skin lesion classification using EfficientNet-B4 on the ISIC 2019 dataset. Ablation study across architectures comparing transfer learning strategies, augmentation pipelines, and fine-tuning approaches. Run on UVA's Rivanna HPC cluster.
PyTorch EfficientNet-B4 Transfer Learning HPC
LLM / RAG
Local NotebookLM Alternative
A free, offline alternative to NotebookLM that runs entirely on Apple Silicon. Chat with your PDFs using a local LLM via Ollama and RAG. Built with LangChain, ChromaDB, and Streamlit. Documented as a Medium article.
LangChain ChromaDB Ollama Streamlit
Agentic AI
Automated Resume Tailor
Dockerized web app that tailors resumes to job descriptions using LLMs. Supports OpenAI, Anthropic, and local Ollama backends via LiteLLM. Built as a portfolio piece demonstrating agentic workflow design and multi-provider LLM routing.
LiteLLM Streamlit Docker Agents
NLP
Seq2Seq Neural Machine Translation
GRU-based encoder-decoder architecture for neural machine translation. Implements attention mechanisms, teacher forcing, and beam search decoding. Comparative analysis of GRU vs LSTM performance on translation benchmarks.
PyTorch GRU Attention NLP

Blog & talks

Mar
2026
From OMOP Pipelines to ClinicalBERT: What EHR Data Engineers Need to Know About Clinical NLP
A practitioner's guide to transformer-based clinical NLP — from why generic BERT fails on clinical text to what fine-tuning actually looks like in a hospital environment.
NLP Healthcare Machine Learning
Feb
2025
I Built a Free, Offline Alternative to NotebookLM — Here's How
How I set up a local LLM on Apple Silicon and built a RAG-powered PDF chat app that runs entirely on my Mac Mini using Ollama, LangChain, and Streamlit.
RAG Ollama LLMs
Mar
2026
NLP for Clinical Text: Clinical BERT-Style Models & LLM Fine-Tuning (YouTube)
Video walkthrough of transformer-based clinical NLP for a Healthcare Data Science course — covering the attention mechanism, the Clinical BERT family, and fine-tuning strategies in practice.
Video ClinicalBERT Healthcare

Experience

Dec 2025 – Present
XponentL Data (Genpact)
Remote
Software Engineer, AI/ML
  • Build and maintain Databricks + Spark pipelines transforming large-scale EHR data into OMOP-aligned formats
  • Develop Python/PySpark transformations for feature engineering, data quality checks, and analytics enablement
  • Support AI-assisted automapping workflows to standardize heterogeneous healthcare schemas at scale
Apr 2025 – Jan 2026
Audible Sight
Remote · Volunteer
AI Engineer
  • Automated Azure-based data processing pipelines for high-dimensional image datasets used in computer vision training
  • Evaluated and benchmarked large language models for generating audio descriptions for visually impaired users
  • Achieved 20% improvement in scene detection accuracy through model optimization
Sep 2021 – Dec 2025
Invictus Capital Partners
Washington, DC
Data Analyst
  • Engineered Spark + Databricks ETL pipelines for multi-terabyte financial datasets supporting analytics and modeling
  • Built Python-based validation and reconciliation workflows, reducing manual processing time by 40%
  • Developed KPI dashboards and reporting frameworks for portfolio monitoring and risk analysis

Let's
connect

Open to roles in AI/ML engineering, applied research, and data science. Graduating August 2026 — available for full-time positions. Willing to relocate to Bay Area, NYC, or Seattle.