Curriculum Vitae
J Justin Donaldson, Ph.D.
[email protected] | jjd.io | cv | linkedin | github
Skills
- Large Language Model Research and Applications
- Data Science and Advanced Machine Learning
- Modeling, Model Optimization, Time Series, Anomaly Detection
- Advanced ML/Web Visualization (Embedding/ML/SVG Related)
- Interaction Design, Prototyping, and Design Theory
- Customer Service Analysis
- Programming Language Design
- Docker, Python, R, + most core web technologies and languages.
Experience
Hushh.ai Cofounder 2024-
Leadership and Engineering Highlights:
- Developed strategies for on-device hybrid LLM search and modeling.
- Coordinated and managed a large team of R&D interns.
Salesforce.com Principal Data Scientist/Engineer Service Cloud 2014-2023
Data Science and Engineering Highlights:
- First Senior Search Datascientist.
- Migrated the core search infrastructure to model-trained ranking coefficients from their previous ad hoc system of boosts and filters.
- Developed and patented deep learning models for predicting integration test failures, and error assignments based on code changes.
- Co-founding member of Service Cloud Datascience team.
- Part of a small team of data scientists tasked with creating an ML education program for all Salesforce engineers.
- Produced the initial Docker Container configuration and serving code that is used as the basis for all production AWS Sagemaker deployments.
- Co-founding member of Foundational AI Components team that produced re-usable libraries and components for advanced ML services.
- Member of joint Engineering/Research Leadership Team responsible for investigating and developing related GPT-3 technologies.
C-level projects with Trusted Salesforce Partners:
- Wrote a custom job candidate scoring function for CPL, Ireland’s largest recruitment firm.
- Wrote an advanced candidate profiler for Allegis, The fourth largest recruitment firm in the world.
- Wrote an anomaly detection model for Hulu to detect when an isolated event was affecting global customer call volume to an unsustainable degree.
- Worked with Citibank data center team to combine customer financial activity with service interactions in order to drive predictive customer issue resolution.
BigML Co-founder/President 2011-2013
Leadership and Engineering Highlights
- Developed and patented interactive visualizations used to represent and explore trained models and distributions.
- Managed various administrative tasks as President (Insurance, Expenses, Office management, etc.)
- Led development of website, and integration with internal/external APIs.
Education
Indiana University - Bloomington, Indiana - 2011
Ph.D, Informatics - January 2011
My Erdös Number is 4:
Dissertation: “Visualization of music relational information sources for analysis, navigation, and discovery”
Advising Committee : Erik A. Stolterman (Chair), Filippo Menczer, Jon Paolillo, Jeff Bardzell, Donald Byrd, Marc Torrens.
President of Graduate Student Body (School of Informatics)
Best Paper Award (J. Donaldson 2006b)
Indiana University - Bloomington, Indiana - 2006
M.S., Human Computer Interaction Design
Depauw University - Greencastle, Indiana - 2001
B.S., Computer Science (Minor in Mathematics)
Publications
Patents
Justin Donaldson. Real-time visualization of user consumption of media items. US8332406B2
J. Justin Donaldson, Adam Ashenfelter, Francisco Martin, Jos Verwoerd, Jose Antonio Ortega, Charles Parker. Visualization and interaction with compact representations of decision trees.
US20200379951A1
Francisco J. Martin, Adam Ashenfelter, J. Justin Donaldson, Jos Verwoerd, Jose Antonio Ortega, Charles Parker. Evolving parallel system to automatically improve the performance of multiple concurrent tasks on large datasets. US9558036B1
Francisco J. Martin, Oscar Rovira, Jos Verwoerd, Poul Petersen, Charles Parker, Jose Antonio Ortega, Beatriz Garcia, J. Justin Donaldson, Antonio Blasco, Adam Ashenfelter. Predictive modeling and data analysis in a secure shared system. US20170140302A1
J. Justin Donaldson, Benjamin Busjaeger, Siddharth Rajaram, Berk Coker, Hormoz Tarevern. Machine learning based ranking of test cases for software development. US10474562B2
Justin Donaldson. Personal music recommendation mapping WO2008051882A3
J. Justin Donaldson, Hormoz Tarevern, Sadiya Hameed, Siddharth Srivastava, Feifei Jiang. Error assignment for computer programs. US10409667B2
Zachary Alexander, Scott Thurston Rickard, Jr., Clifford Z. Huang, J. Justin Donaldson. Accounting for positional bias in a document retrieval system using machine learning. US10565265B2
Scott Thurston Rickard, Jr., Clifford Z. Huang, J. Justin Donaldson. Adjusting feature weights for ranking entity based search results. US20180052853A1
Pro Bono
University of Washington - Instructor/Contributor
Co-Instructor for UW 410 Advanced Machine Learning.
(Slides Available: https://jdonaldson.github.io/uw-mlearn410/)Member of UW Advisory Board for their school of Professional and Continuing Education
Haxe Foundation - Contributor/Member
Developed the Lua target for the Haxe language blog announcement | hacker news fp
Sponsored Summer-of-Code for Haxe programmers through mentorship and financial contributions.
UCLA Graduate School - Data Science Advisor/Mentor
- Co-hosted a graduate student Data Science competition in collaboration with Boston Day Academy (Slides Available : https://is.gd/qnji8A)
© Copyright 2024 Justin Donaldson. Except where otherwise noted, all rights reserved. The views and opinions on this website are my own and do not represent my current or former employers.