Experience

Octave Bioscience

• Lead Data Scientist

  • Nov 2021 - Present

  TL;DR Leverage data & algorithms to give care teams actionable & transparent insights to help people with neurodegenerative diseases access expert proactive care regardless of their geographic location. I am developing population heath solutions for neurodegenerative disease management (i.e. making inferences about a patient based on data like medical claims to facilitate management of a population at scale). This includes designing & building data pipelines to create clinically relevant features, applying granular subject matter expert knowledge together with data-driven insights/models to translate clinical intuition + medical transaction data into informative & actionable information about a population and individuals. I also do a fair amount of financial modeling to identify where there are opportunities to improve care so healthcare dollars are used more proactively to improve outcomes. Show less



Rally Health

• United States
• Health, Wellness & Fitness
• 200 - 300 Employee

• Senior Data Scientist

  • Jan 2019 - Nov 2021

  Optum Digital (Formerly Rally Health) I worked on various digital health products, primarily on tools to support cost-efficient decision making for members navigating the healthcare system. I worked closely with product managers on our "Find & Price Care" solution (e.g. price transparency tool on uhc.com) and on data-driven incentive design. One of my projects applying retail-sector modeling approaches to the healthcare space for "Shoppable" care (e.g. routine MRIs or labs) took first place in an internal product competition for funding. I also collaborated with VP-level stakeholders on short-term projects like simulating distributions of financial outcomes in member populations to support contract negotiations. Show less



Diffusion Imaging in Python

• United States
• Software Development
• 1 - 100 Employee

• Developer

  • Mar 2018 - Jan 2019

  The python open-source community got me started in software development, supporting me as I progressed from a first-time contributor to developer in the space of a few years. It can be intimidating to put your code out there, but it's an awesome way to learn, and this community is amazing! See the link, below, for my two cents on getting involved, and feel free to reach out, especially if you're considering a first-time contribution.

• Contributor

  • Jul 2016 - Mar 2018



UCSF Memory and Aging Center

• United States
• Hospitals and Health Care
• 100 - 200 Employee

• Postdoctoral Research Fellow

  • Aug 2017 - Jan 2019

  As a Bioengineering/Neurolinguistics Postdoctoral Fellow at the UCSF Memory and Aging Center and UCSF Dyslexia Center I developed multimodal image processing methods to study Neurodevelopmental (Dyslexia) and Neurodegenerative (Primary Progressive Aphasia) Disorders and applied interpretable machine learning methods to the diagnosis of Neurolinguistic Disorders using Neuropsychology & Audio data in collaboration with domain experts. I conceived of, designed, and implemented an automatic tractography* model segmentation algorithm that replaced the tedious manual work of a specialized Neuroradiologist. *Tractography is essentially a Markov Chain Monte Carlo (MCMC) sampling method used to create probabilistic models of brain connections built upon physics models of water diffusion, measured by specialized Magnetic Resonance Imaging (MRI) techniques, constrained by Neuroanatomical assumptions at the micro-, meso-, and macro-scale. I worked on the open-source library Diffusion Imaging in Python (DIPY), which implements tools to work with this class of methods, developing application-based techniques for the macro-scale problem of constraining these models to Neuroanatomically feasible connections. Projects - Characterizing AV1451 Binding (Tau-PET) in Nonfluent Variant Primary Progressive Aphasia White Matter Structures - Developing Automatic Approach to White Matter Anatomical Modeling with Diffusion MRI Fiber Tracking - Model White Matter Structure in Children with Dyslexia to Explore Neurodevelopmental Features - Classify Neurolinguistics Disorders based on audio data Show less



Insight Data Science

• United States
• Higher Education
• 1 - 100 Employee

• Insight Health Data Science Fellow

  • Sep 2018 - Dec 2018

  As an Insight Data Science Fellow in the health space, I created a screening tool for early signs of Neurodegeneration based on audio files of a patient reading a standard passage. I leveraged the Nonlinear Dynamic Timewarp algorithm in an off-label way. This algorithm is typically used in automatic speech recognition to map variable spoken audio clips (e.g. "hello", "heellllloooo") to the same word. I mapped the audio files to a template of myself speaking the passage, and then calculated the amount of nonlinear warp timestamp-by-timestamp required to make the passages fit together. I used these nonlinear timestamp features (as well as a linear scaling factor) as the input to my model. The resulting important features left after regularization gave a timestamp-specific marker which mapped to parts of the passage that neurodegenerative patients struggled with. A speech-language pathologist could then interpret what the model had used to distinguish neurodegenerative patients from healthy controls. This approach was inspired by the typical process to look at visually apparent degeneration in the brain, in which we register images to a template and look at the jacobian determinant of the warp field to identify which voxels (3D pixels) represent atrophied brain tissue. Show less



University of California, San Francisco

• United States
• Higher Education
• 700 & Above Employee

• PhD Candidate UC-Berkeley/UCSF Joint Program in Bioengineering

  • Aug 2012 - Jun 2017

  I completed my doctorate in Bioengineering with the dissertation "Diffusion MR Image Processing Tools for Reliable Fiber Tracking* Analyses: Neurosurgery and Radiation Oncology Applications" at UCSF. Traditional image processing & tractography* modeling tools often fail when an extreme pathology is involved, so I developed novel approaches to study how the brain works when it is injured by a tumor or treatment, enabling tools to maximize outcome for Neurosurgery/Radiation Therapy patients. My work focused on creating maps of the brain that give clinicians a better sense of how to protect a patient’s quality-of-life (ex: movement, language, vision) when they must damage the brain to treat a pathology. I was responsible for supervising multiple research projects by clinical residents and fellows, and for the technical administration of a clinical research service providing brain circuit modeling for brain tumor surgeries, as well as training many clinical and research stakeholders in tractography methods and applications. I regularly gave research & educational talks including an invited Psychiatry Grand Rounds lecture on Network Theory & Brain Connectivity Research (Queens University, ON), and an invited talk at BrainHack Global (IU). I contributed to a chapter on "Lesion-behavior awake mapping with direct cortical and subcortical stimulation" in the book "Lesion-to-Symptom Mapping: Principles and Tools" (in press). I also worked on projects in Multiple Sclerosis, Epilepsy, Depression/Anxiety, and Parkinson’s Disease. *Fiber Tracking is essentially a Markov Chain Monte Carlo (MCMC) sampling method used to create probabilistic models of brain connections built upon physics models of water diffusion, measured by specialized MRI techniques. I worked on the open-source library Diffusion Imaging in Python (DIPY) and developed application-based techniques for the macro-scale problem of constraining these models to Neuroanatomically feasible connections. Show less



University of California, Berkeley

• United States
• Higher Education
• 700 & Above Employee

• Graduate Student Instructor, Bioengineering Capstone Senior Design

  • Aug 2014 - Dec 2014

  As a graduate student instructor, I supervised senior bioengineering undergraduate capstone design projects. I was responsible for guidance, grading, interfacing with speakers from industry, and teaching. As a graduate student instructor, I supervised senior bioengineering undergraduate capstone design projects. I was responsible for guidance, grading, interfacing with speakers from industry, and teaching.


U.S. Army Research Laboratory
• United States
• Armed Forces
• 500 - 600 Employee
• Intern/Technician

  • May 2009 - Aug 2012


University of Maryland

• United States
• Higher Education
• 700 & Above Employee

• Undergraduate Researcher, Muro Lab

  • Sep 2011 - May 2012

  Gained wet bench research experience through an Engineering Honors Program thesis investigating receptors for targeted drug delivery. Gained wet bench research experience through an Engineering Honors Program thesis investigating receptors for targeted drug delivery.



Engineers Without Borders USA

• United States
• Non-profit Organizations
• 100 - 200 Employee

• Peru Water Sanitation Project

  • Sep 2008 - May 2011

  • Served as subgroup leader of community awareness team and member of design team (2008 – 2009)• Deployed to Compone, Peru: implemented grassroots community sanitation awareness plan in the schools and participated in construction of water sanitation system (2009)• Served as member of design team in the capacity of resource (2010 – 2011)

• Treasurer

  • May 2009 - May 2010

  • Responsible for a large budget that fully supported 3-4 projects simultaneously• Developed budget proposal to gain funding from the University Student-Government Association• Presented to academic departments to solicit funds• Oversaw 3 – 4 engineering projects all over the world and an organization of over 100 students, professionals, and faculty as a member of the Executive Board


Vanasse Hangen Brustlin
• United States
• Medical Practices
• Intern

  • Jun 2007 - Aug 2007