Livingston, Montana, United States
I’m a builder who thrives at the intersection of technology, people, and execution. I seem to have a knack (or curse) for gravitating toward complex, high-stakes problems, and I take a lot of pride in turning them into working systems that deliver real value. My career has combined deep technical work with hands-on entrepreneurial experience, including several years running an international wellness business. That chapter tested everything: my ability to lead without structure, adapt in real time, and build lasting relationships with people from wildly different walks of life. It gave me a toolkit that goes far beyond code, built on resilience, clarity, and the confidence to figure things out when there's no obvious path. In recent years, I’ve been back in tech, working with startups to design and deploy AI pipelines that process massive volumes of media, applying computer vision techniques to extract structured, meaningful insights from raw video, images, and audio. My focus spans the full lifecycle: evaluating and integrating the right models, shaping the data through thoughtful preprocessing and postprocessing, and building the infrastructure and ML operations to run these systems reliably and efficiently at scale. A big part of the work involves making practical decisions weighing accuracy, speed, and cost to ensure the systems aren’t just technically sound, but also sustainable and aligned with business goals. I’m at my best while working with people who care deeply, communicate openly, bring both humility and conviction to solving hard problems, and of course have some fun along the way.
- Screening, trial evaluation, and onboarding operations for Sunday's Memory Developer program — the human data collection workforce behind Memo's training pipeline - Building internal automation tools to streamline recruiting and ops workflows as we grow to 2000 MemDevs
Using advanced capture systems to record real-world actions and build foundational training data for Sunday's home robots
I remain on call to help the team navigate the more intricate corners of the pipelines I originally built. - Provide ad-hoc architecture and debugging guidance when tricky edge cases surface. - Offer light mentorship / code reviews as bandwidth allows.
I led the design, build, and operation of Egoli’s computer-vision platform, which now processes petabytes of unstructured video and image data. - Architected cloud-native, serverless pipelines handling image classification, face recognition, OCR, object detection, and audio metadata extraction—all orchestrated with Step Functions + Lambda + ECR containers. - Owned model lifecycle: evaluation, preprocessing, post-processing, and periodic retraining; balanced speed, accuracy, and AWS billing constraints. - Built reproducible environments with Docker & CloudFormation; championed infra-as-code and CI/CD. - Integrated GenAI vision APIs where they beat classical CV, but kept traditional models where they still outperformed—always measuring ROI, not hype. - SQL-first data stewardship: designed schemas, wrote 100-plus optimization scripts, and kept RDS/MySQL humming during media-ingest surges. - Reported directly to leadership on technical trade-offs, cost forecasts, and product impact.
Founding engineer on Egoli’s start-up team, turning research ideas into production-grade proofs of concept - Prototyped and benchmarked state-of-the-art CV models for facial recognition, OCR, scene detection, audio transcription, object detection, and classification. - Ran large-scale experiments—thousands of GPU hours—to compare model accuracy, inference cost, and latency under real traffic. - Developed statistical tooling for error analysis, informing later pipeline QA gates. - Architected database schema and implemented data science code into batch production.
I lead MLOps at MyVisionLabs.ai, focusing on AI-powered sports coaching. I design and build the cloud infrastructure for our real-time computer vision analysis, enabling athletes to receive instant feedback on their form and technique. I work closely with both the data science research and UX teams, bridging the gap to ensure technical solutions align with user needs. My responsibilities include: - Designing scalable, low-latency, and cost-optimized cloud infrastructure for a robust set of computer vision models - Building resilient ML pipelines for real-time video processing - Managing the integration of data science research code into end-to-end production deployments
Los Angeles Yoga Instructor (November 2016-March 2020): - Taught between 6-12 yoga classes per week at top studios & gyms in Santa Monica & West LA - Regular private sessions with wide spectrum of clients ranging from from families introducing their children to yoga, athletes facing injury rehabilitation, celebrities who needed a break from the insanity. Travelling Yoga Instructor (April 2013 – September 2016): - Settled in Brisbane, Australia from August 2013-May 2014 & Wellington, New Zealand from September 2014-August 2015, teaching 15/20 yoga classes/week across multiple yoga studios, gyms, private & corporate venues - When not settled, toured for total of 20 months across Oceania, Asia & Europe, leading retreats and Acroyoga/Inversion workshops Studio Owner - Yoga Loft Tampa (October 2011-April 2013) - Presided over all operations & marketing of a 3-month old yoga studio, designed & created website www.yogalofttampa.com & performed SEO to make Google’s 1st page for query: “yoga in tampa” - Taught between 15-20 classes per week, simultaneously creating “Saturday Yoga in Ybor” community event which consistently draws 50-75 students for donation yoga in the Ybor City Museum Gardens - Monthly attendance & revenue doubled in my first month upon arrival and grew 5X by end of first year
- Reintroduced myself to the big changes in the tech world with this part-time gig as I built up my yoga credentials in LA - Earned Certifications in the software offerings of our IT partners (Splunk, NetScout, Docker, AWS) and led the tech side of sales discussions with our Enterprise clients (NBC, OceanX)