I am a Staff Machine Learning Engineer at Apple, where I build large-scale machine learning systems for search and data-centric artificial intelligence. My early work focused on AppleBot - Apple's web crawler, among the top ten AI crawlers on the web - and on information extraction systems that power Spotlight and Siri. Since 2023, I have been leading multiple efforts in pre-training data quality and curation for the foundation models behind Apple Intelligence.

Before Apple, I was a Machine Learning Researcher at NASA's Jet Propulsion Laboratory, where I worked on the DARPA Memex program under the supervision of Dr. Chris Mattmann, building search technologies for the deep and dark web. I also contributed to the Mars Target Encyclopedia, applying text mining to extract scientific findings from planetary science literature, and to Polar Deep Insights, an EarthCube-funded project surfacing scientific data from polar repositories.

I received my master's degree from the University of Southern California, supported by a Dean's Master's Fellowship. There, I co-invented Sparkler, a distributed web crawler built on Apache Spark with an Apache Solr backend, and contributed to other Apache projects such as Nutch - work that led to my election as a member of the Apache Software Foundation.


Highlights

EACL 2026 [2026] Paper Beyond a Single Extractor: Re-thinking HTML-to-Text Extraction for LLM Pre-training accepted at EACL 2026
Apple Foundation Models Architecture [2025] Launched new versions of Apple's On-Device and Server Foundation Language Models supporting several new capabilities and languages
Apple Intelligence [2024] Apple Intelligence launched at WWDC; contributed to pre-training data quality and curation for the foundation models. Press: CNBC, TechCrunch, The Verge, NYT
MM1 [2024] Our paper on MM1, Apple's multimodal model family, appeared at ECCV 2024. Featured in Wired, VentureBeat, Nasdaq
Apple [2017] Joined Apple as part of the Search team
Sparkler talk at Spark Summit [2017] Presented Sparkler at Spark Summit East (video, slides); also spoke at ApacheCon North America and ApacheCon Big Data Europe
Podcast [2017] Featured on the Science and Supercomputers podcast — "When Data's Deep, Dark Places Need to be Illuminated" — discussing how we used the TACC Wrangler supercomputer to combat human trafficking through deep web analysis
Polar Deep Insights [2017] Presented Polar Deep Insights at the EarthCube All Hands Meeting — mining scientific data from polar repositories using Sparkler and NSF XSEDE supercomputing resources [NSF award]
Apache Software Foundation [2017] Elected as a member of the Apache Software Foundation [announcement]
DARPA Memex [2016] Joined NASA JPL as part of the DARPA Memex program to build search technologies for the deep and dark web

Publications

See Google Scholar for a complete list, or view the Global Author Citation Map.

Beyond a Single Extractor: Re-thinking HTML-to-Text Extraction for LLM Pretraining
J. Li, J. P. Gardner, D. Kang, F. Shi, K. Singh, et al. EACL 2026
[paper] [eacl]
Apple Intelligence Foundation Language Models: Tech Report 2025
E. Li, A. B. L. Larsen, ...K. Singh... et al. Apple, 2025
[paper] [arxiv]
MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training
B. McKinzie, Z. Gan, ...K. Singh... et al. ECCV 2024
[paper] [arxiv] [wired] [venturebeat]
Apple Intelligence Foundation Language Models
T. Gunter, Z. Wang, ...K. Singh... et al. Apple ML Research, 2024
[paper] [arxiv] [product] [wikipedia]
Mars Target Encyclopedia: Rock and Soil Composition Extracted from the Literature
K. Wagstaff, R. Francis, T. Gowda, Y. Lu, E. Riloff, K. Singh, N. Lanza. AAAI, 2018
[paper] [project]
An Automated Approach for Information and Referral of Social Services Using Machine Learning
M. Sharan, N. K. Ottilingam, C. A. Mattmann, K. Singh, et al. IEEE IRI, 2017
[paper]

Open Source

Sparkler — Co-Inventor
A distributed web crawler built on Apache Spark that enables large-scale crawling with integrated NLP and ML capabilities
Apache Nutch — Committer & PMC Member
A highly extensible and scalable open-source web crawler built on Apache Hadoop. Used widely in production search and data-mining pipelines
Apache DRAT — Committer & PMC Member
A Distributed Release Audit Tool that automates license header checking and code compliance analysis across large codebases

Recognition & Engagement