Yale University, Dept. of Biomedical Informatics &
Data Science
Senior Software Engineer
June 2025 -
Current
New Haven, CT
Implemented edge bundling and visualization for
MedViz, an application that maps the semantic space of
biomedical literature.
Created an agentic workflow for generating knowledge
graphs for drug, gene and disease relationships.
Engineered agentic ETL pipelines for extracting
biomedical datasets as part of
DataMed
— an NIH-backed index enabling researchers to search,
identify, and retrieve data for model training and
scientific research. The pipeline features a
multi-agent, self-healing execution loop that
dynamically generates Dagster assets and manages
end-to-end pipeline orchestration. To date, the system
has indexed metadata for over 6.5 million datasets
spanning 241 NIH repositories.
Designed and built a data pipeline and API for the NIH
Data Sharing Index (S-index) Challenge, aggregating
repository datasets and authors to generate identifiers
for AI-driven publication matching and enabling search
via a dedicated
S-index application.
Pydantic-ai
MCP
Three.js
Vue
Dagster
Python
FastAPI
Postgres
Elasticsearch
Trinnex
Senior Full-Stack Developer
March 2023 - June 2025
Boston, MA
Worked on
LeadCAST, a platform used by water companies to visualize,
track and schedule the replacement of water lines.
Led the development of the replacement module, which
maps and visualizes lead pipes, predicts lead statuses
and tracks waterline inspections and replacements.
Created an AI powered workflow for lead status
predictions. This accelerated customer lead replacement
programs by reducing the number of required inspections
from thousands to hundreds.
Implemented an event driven architecture to create
export services that generate real time reports of
replacements and water line inventories.
Next.js
React
Node
Express
.NET
Pubsub
GCP
Postgres
MongoDB
Infinitive Inc.
Senior Developer
April 2021 - December 2022
Ashburn, VA
For our client
Unite Us, a healthcare platform, revamped the provider
onboarding services workflow and reduced the process
down from days to hours. This had a significant impact
by increasing the new customer acquisition rate by 30%.
Created an organization deduplication service for the
ETL pipeline which eliminated the need for manual
review.
Developed a pose detection neural network to classify
golf swings using Pytorch and Sagemaker.
React
Go
Ruby
Python
Neo4j
Postgres
gRPC
AWS
Artemis Consulting
Python Developer
June 2019 - April 2021
Washington DC
Contractor for the
Library of Congress’ Chronicling America
project: a digital collection of over 2.6 million
newspaper scans curated by libraries and universities
throughout the country.
Implemented the batch download service leveraged by LOC
archivers to rapidly download collections of data
including PDF scans, OCR text and categorical data.
Created a D3.js geo-visualization of newspaper
collections by library/awardee. Users can filter by an
array of facets and see their results displayed on the
map.
Extended a OCR based highlighting microservice to
include non-English characters for passages in languages
such as Hebrew and Arabic.
Python
Django
Flask
D3.js
Redis
Solr
AWS
Duke University, Curtarolo Materials Lab
Research Assistant
April 2015 - June 2019
Durham, NC
Maintained, deployed and developed applications for
AFLOW, a materials analysis platform and database of over
3.9 million material structures with 817 million
properties.
Co-developed predictive models for the electronic and
thermal properties of materials. Modeling techniques
include graph convolutional neural networks, XGBoost and
traditional methods.
Created
AFLOW ML, a Python library and REST API of predictive models
for superconductivity, and metal and insulator
classification. Users include researchers at Hyundai,
UNC Chapel Hill and The Office of Naval Research.
Pytorch
Tensorflow
Pandas
Scikit-learn
Python
C++
CUDA
Yale University, Schroers Lab
Freelance Developer
December 2014 - April 2015
New Haven, CT
Developed the prototype for
The Materials Atlas Project, an application that allows research groups store and
visualize characterization data of metallic glasses.
Visualizations using D3.js and Three.js include ternary
plots, composition gradients and interactive sample
library overviews which display optical images when
selecting a specific point.
Python
Django
D3.js
MySQL
The Research Center on Computing & Society, Southern
Connecticut State University
Web Developer
October 2013 - December 2014
New Haven, CT
Mantained the Wordpress site for the The Research Center
on Computing & Society. Duties included updating pages,
generating content and maintaining the site.
Led a full site migration to a new server, upgraded
WordPress, and resolved all plugin versioning and
compatibility issues.