Daniel Whitenack
Also known as Data Dan
Daniel Whitenack is an AI builder whose work spans low-resource language technology and enterprise AI governance, and he cohosts Practical AI. He earned a PhD in physics at Purdue in 2012 and a B.S. in engineering physics from the Colorado School of Mines. His engineering work included GPU-accelerated TensorFlow on Kubernetes and maintaining the Go kernel for Jupyter. At SIL International, he worked on speech synthesis and cross-lingual translation in 2020; his 2022 and 2024 papers developed multimodal language-model training and a dictionary tool for low-resource languages. Prediction Guard raised $3.7 million in 2025 and released a Docker Hub agent kit in September 2026; as of October 2026, he is its founder and CEO.
Timeline
-
2012
A Purdue physics PhD
He earns a PhD in physics from Purdue in 2012 and holds a B.S. in engineering physics from the Colorado School of Mines.
-
2017
A Go kernel for Jupyter
In 2017, he writes about GPU-accelerated TensorFlow on Kubernetes for O’Reilly. He maintains the Go kernel for Jupyter and helps organize open-source data-science contributions.
-
2018
Cohosting Practical AI
The Practical AI podcast opens in July 2018 with Daniel Whitenack and Chris Benson discussing artificial intelligence, machine learning, and data science and their plans as hosts.
-
2020
Speech and translation tools
At Project Voice in 2020, he presents a workshop on natural-sounding speech synthesis for local languages and talks on voice applications for emerging markets and AI for local languages as a member of SIL’s Innovation, Development & Experimentation department. He uses cross-lingual embeddings with SIL colleagues to construct “wash your hands” phrases in 544 languages.
-
2022
Phone-based language models
In 2022, he and Colin Leong propose a multimodal method for training language models from whatever text or audio is available for a language, with experiments on Swahili and Kinyarwanda.
-
2024
The GUIDE dictionary tool
In 2024, he coauthors GUIDE, a graph-neural-network tool that creates semantic-domain dictionaries for low-resource languages from seed dictionaries and parallel Bible translations.
-
2025
A $3.7 million seed round
In May 2025, Prediction Guard raises a $3.7 million seed round to expand product adoption in regulated industries.
-
As of October 2026
Leading Prediction Guard
As of October 2026, he is founder and CEO of Prediction Guard and cohosts Practical AI. In September 2026, the company releases an agent control kit on Docker Hub for governed agents, with policy enforcement and observability.
Career
- NowPrediction GuardFounder and CEO
- 2020SIL InternationalMember
- 2020–2022SIL Global
Education
- 2012Purdue UniversityPhD
- Colorado School of MinesB.S.
Most-cited works
- Participatory Research for Low-resourced Machine Translation: A Case Study in African Languages
- Container-based bioinformatics with Pachyderm
- Masakhane -- Machine Translation For Africa
- Density Functional Resonance Theory of Unbound Electronic Systems
- Effects of magnetic dipole-dipole interactions in atomic Bose-Einstein condensates with tunable s -wave interactions
Sources and credits
Photo: We have no published photo of Daniel Whitenack yet. Send us a free-licensed photo.
- Solving Social Challenges | Purdue Daniels School of Business Journal
- Leadership | ExploreGod
- Daniel Whitenack – O’Reilly Author(s) – O’Reilly
- Meet your Practical AI hosts | Practical AI
- Advancements in voice technology open new doors for local language communities | SIL Global
- How I’m using AI to translate ‘wash your hands’ in 500 languages | Opensource.com
- Phone-ing it in: Towards Flexible Multi-Modal Language Model Training by Phonetic Representations of Data | ACL Anthology
- GUIDE: Creating Semantic Domain Dictionaries for Low-Resource Languages | ACL Anthology
- $3.7M in Seed Funding for Prediction Guard to Accelerate Growth
- About | Prediction Guard
- Prediction Guard Joins Docker Verified Publisher Program and Releases Agent Control Kit on Docker Hub
