Justin Reppert
Machine Learning Research Engineer at Elicit
Austin, Texas, United States
Summary
👤
Senior🎓
Top SchoolJustin Reppert is a Machine Learning Research Engineer with a decade of experience building production-grade NLP systems and leading ML efforts at Elicit, where he oversees ML infrastructure, inference, and multinode training pipelines up to 70B models. With a PhD in the philosophy of mathematics, he combines rigorous, conceptual thinking with practical engineering—specializing in large-scale hybrid semantic search and applying language models to complex, high-cost-feedback reasoning tasks. He runs and ships end-to-end systems (Knative on k8s, Triton/vLLM, FAISS-backed retrieval) and prioritizes trustworthiness, interpretability, and controllability in deployed NLP tools. Justin enjoys the pace and ownership of startups and has a habit of creating durable side projects—like a continuously running music-similarity search service he built on a tiny DigitalOcean droplet that has required almost no maintenance since 2021.
10 years of coding experience
7 years of employment as a software developer
Bachelor of Arts, Bachelor of Arts at Wheaton CollegeDoctor of Philosophy - PhD, Doctor of Philosophy - PhD at Fordham University Graduate School of Arts and Sciences
Github Skills (47)
debugging10
language-model10
python10
tensorflowjs10
music10
contrastive-learning10
fastapi9
scikit9
react8
self-supervised-learning8
metric-learning8
image-retrieval8
lenskit8
pytorch8
paths7
Programming languages (6)
Github contributions (5)
reppertj/philosophical-graphiti
Jun 2020 - Apr 2021
A webapp for exploring and visualizing subgraphs induced on the shortest paths between two nodes. Originally used with data from the Stanford Encyclopedia of Philosophy.
Contributions:11 commits, 4 PRs, 23 pushes in 9 months
stanfordnodesencyclopediapathsvisualizing
reppertj/earworm
Feb 2021 - May 2021
Search for music by sonic similarity
Contributions:175 commits, 26 pushes, 1 branch in 2 months
embeddingsmusicreacttensorflowjspytorch
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.