Dean Lebaron Email & Phone Number

Salt Lake City, Ut, Us

Utah, United States

Recently, I made an internal move to one of Lucid's ML/AI teams. I'm currently working in a mostly ops role doing things like architecting AWS infrastructure, load testing models, optimizing ML run times, and similar. Before this I spent a few months on a Scala backend team working on billing integration.

For a few brief, golden months I was able to be a technical product manager. This included- Roadmapping new features- Prioritizing ongoing ML work vs new services and systems.- Educating stakeholders about ML capabilities and cooperatively exploring where ML could offer solutions. - Managing risk and expectations around various ML backed services. - Monitoring and reporting metrics and keeping stakeholders informed of progress.

Lehi, Utah, United States

I provide operational support to our ML teams. What this looks like on a day to day is deploying and maintaining model serving apps, evaluating and optimizing production ML pipelines for speed, cost and throughput, supporting data scientists in data set collection and analysis, etc.

Lehi, Ut

I work at Geomni (a subsidiary of Verisk) on the Geomni Property Essentials team researching and developing computer vision software for the automation of tasks such as annotating and segmenting images, modeling structures, and detecting property features. I have been responsible for deploying our machine learning solutions to AWS and architecting a cloud-based image processing pipeline to produce our data at scale.I am currently involved in leading the machine learning development on the Triage product. This includes defining technical requirements, planning and roadmapping intermediate and final solutions, and orchestrating development across multiple teams. I also frequently discuss and refine features with business stakeholders.

Neural Networks And Machine Learning Lab - Advisor Tony Martinez

I am currently working with various clustering algorithms using some active learning/semi-supervised aspects to improve clustering of handwritten images. Specifically, I am focusing on assigning confidence scores to individual instances. This is accomplished through treating the original clustering as a sort of "pseudo-labeling" and the training supervised learners on the clustered data. The scores from the learners can then be used as a measure of confidence in a given instance.

Orem Ut

- Worked on back-end Java web services and domain logic with Jersey, Spring, Hibernate, and other frameworks.- Wrote automated web tests using Selenium

- Assisted students in CS 142 - Intro to Programming and CS 235 - Data Structures.- Taught regular help sessions and explained basic concepts- Assisted students in debugging and designing assigned projects

- Taught regular help sessions covering topics such as dynamic programming, simplex method, branch and bound algorithms, divide and conquer algorithms, etc.- Assisted students in debugging and designing assigned projects

Master'S Degree, Computer Science

Activities and Societies: ACM

Bachelor'S Degree, Genetics And Biotechnology

Activities and Societies: ACM