Bruce Bugbee

Bruce Bugbee

Hi! I'm a Colorado-based data scientist who's worked across a bunch of different fields. I like tinkering with AI, hacking things, filling whiteboards with half-baked ideas, getting way too competitive at NYT Connections, training Brazilian Jiu Jitsu, playing D&D, anthropomorphizing my dog Remy, and chasing my two kids around—usually while coming to terms with the fact that my favorite albums are now 20+ years old.

Recent Post

I Finally Built My Website

2026-01-13

Well I guess I have a website now. This is very 2000s of me.

I've been an avid Hard Fork podcast listener for the last few years. On a recent episode the hosts had a segment where they detailed their most recent vibe coding explorations using Claude Code. Interestingly enough, the first thing each of them did was to use Claude Code to fully build out their own personal websites, replacing services like Squarespace in the process.

Continue reading

Career Highlights

Airtable

2025 – Present

Brought in to start up the machine learning function within Airtable's Data team, building advanced solutions using AI and ML capabilities. Focused on establishing the foundational infrastructure needed to productionize ML and LLM systems at scale—from model training and experimentation platforms to batch inference pipelines and serving infrastructure.

Key achievements:

  • Built end-to-end ML Ops infrastructure: model training, experimentation, serving, and batch LLM inference capabilities
  • Shipped first production ML model improving marketing lead-scoring accuracy by ~50%, directly impacting pipeline quality

Amazon

2024 – 2025

Led science efforts around Amazon's incrementality measurement platform, with emphasis on developing improved causal attribution models for advertising across channels on Amazon. Focused on helping advertisers understand true campaign effectiveness through rigorous causal inference and experimentation.

Key achievements:

  • Led cross-disciplinary science team owning the advertiser-facing RCT experimentation platform
  • Established technical standards and best practices for causal ML models powering Multi-Touch Attribution products
  • Developed strategic roadmap for extending causal inference capabilities across international marketplaces
  • Built evaluation frameworks for measuring attribution model quality and business impact

CaliberMind

2022 – 2024

Brought on after CaliberMind's Series A to handle all things data in an early-stage startup environment. The core challenge: building unified data abstractions that could integrate each B2B customer's unique data structure into a common analytical framework. Led everything from data connection strategy and metric definition to attribution modeling and analytics infrastructure.

Key achievements:

  • Designed analytical abstractions and unified data layers that allowed diverse B2B customer data to integrate seamlessly into the attribution platform
  • Built ML models for multi-touch attribution that worked across varied customer data structures and marketing touchpoints
  • Established foundational data practices around connection management, metric definition, and data integrity

Oracle

2018 – 2022

I spent nearly four years with the audience modeling group at Oracle Advertising (formerly Oracle Data Cloud), where we built ML systems that turned diverse data sources—online behavior, purchase data, demographics—into 3rd party audiences for advertisers. I was a full-stack data scientist doing everything: model development, large-scale data engineering, data quality alerting, DevOps, the works. The team eventually expanded into contextual classification for open web advertising. Started as a Principal Data Scientist and worked my way up to Director.

Key achievements:

  • Led ML systems for audience targeting and activation across Oracle's advertising products
  • Modernized legacy platforms delivering $400K annual cost savings, ~30% performance lift, and 10× runtime improvements
  • Built shared ML infrastructure enabling reliable model deployment, monitoring, and A/B testing across teams
  • Managed cross-functional data science team while maintaining hands-on technical leadership

National Renewable Energy Lab

2015 – 2018

Got to work on a wide array of data problems across renewable energy fields—energy systems modeling, high-performance computing optimization, transportation analytics, and more. This is where I learned to take complex statistical methods and make them useful for domain experts.

Key research:

  • Applied machine learning and statistical methods to energy system optimization, HPC power prediction, and vehicle telematics analysis
  • Published research on interactive visualization combined with deep learning for real-time energy model exploration
  • Analyzed telematics data from 57,000+ vehicles traveling 210M+ miles to assess truck platooning feasibility

MD Anderson Cancer Center

2014 – 2015

Worked on problems in functional data analysis with applications to cancer medicine and RNA data. First position after finishing my PhD, applying computational statistics to real medical research.

Education

Colorado State University

2008 – 2014

Advisor: Jay Breidt

Advisor: Haonan Wang

Colorado School of Mines

2004 – 2008

Publications

Selected peer-reviewed publications spanning computational statistics, machine learning, energy systems, and data science. Research focuses on translating advanced statistical methods into practical systems.

Statistics & Computational Methods

Bugbee, B.D., Breidt, F.J., & van der Woerd, M.J. (2016)

"Laplace Variational Approximation for Semiparametric Regression in the Presence of Heteroskedastic Errors."

Journal of Computational and Graphical Statistics, 25(1), 225-245.

Bugbee, B.D. (2014)

"Semiparametric Regression in the Presence of Complex Variance Structures Arising from Small Angle X-Ray Scattering Data."

PhD Dissertation, Colorado State University. ProQuest Dissertations & Theses.

Hayne, S.C., Bugbee, B., & Wang, H. (2010)

"Bidder behaviours on eBay: collectibles and commodities."

Electronic Markets, 20, 95-104.

Energy Systems & Machine Learning

Bugbee, B., Bush, B.W., Gruchalla, K., Potter, K., Brunhart-Lupo, N., & Krishnan, V. (2019)

"Enabling immersive engagement in energy system models with deep learning."

Statistical Analysis and Data Mining: The ASA Data Science Journal, 12(4), 325-337.

Potter, K., Brunhart-Lupo, N., Bush, B., Gruchalla, K., Krishnan, V., & Bugbee, B. (2017)

"Coupling Visualization, Simulation, and Deep Learning for Ensemble Steering of Complex Energy Models."

2017 IEEE Workshop on Data Systems for Interactive Analysis (DSIA), Phoenix, AZ, 1-5.

Bugbee, B., Phillips, C., Egan, H., Elmore, R., Gruchalla, K., & Purkayastha, A. (2017)

"Prediction and characterization of application power use in a high-performance computing environment."

Statistical Analysis and Data Mining, 10(3).

Transportation & Telematics

Lammert, M.P., Bugbee, B., Hou, Y., Mack, A., et al. (2018)

"Exploring Telematics Big Data for Truck Platooning Opportunities."

SAE Technical Paper 2018-01-1083. WCX18: SAE World Congress Experience, Detroit, MI.