Leading research on the Mamba state space language model to develop personal AI products that can run locally on conventional computers.

Published video courses on LLMs, GANs, and other generative AI technologies for developers and data scientists worldwide.

- Leading mentorship conversations with recent graduates fromalternative education backgrounds- Sharing advice and best practices with junior level engineers and datascientists who are starting their career- Educating junior level engineers and data scientists about how toelevate themselves in the workplace- Inspiring and uplifting junior level engineers and data scientists tohelp them build confidence in their work

HiCounselor is a career accelerator that provides the coaching and resources job seekers need to land their dream job.As an instructor, I've lead sessions on Product Analytics, SQL, and Interview strategies to help cohorts of college students, recent graduates, and professionals prepare for Data Science Interviews.I've also performed dozens of mock interview sessions with candidates and helped them market their skills by reviewing their resume.I also produced several hours of Data Science and Machine Learning video course content on- Machine Learning- Statistics and Probability- Product Analytics- Interview Strategies and Tips- Distributed Computing- System Design

Parle Innovation is the producer of the eyeVue, a smart AR binocular that enhances the live viewing experience of sports fans and spectators. I worked with the executive team to build the eyeVue product MVP. My work resulted in $10k in gross sales within 6 months of product release.My work included implementing the core functionality of eyeVue smart binocular firmware using Ardunio C++, building out the companion iOS and Android app using Swift, Kotlin, and Java, and making iterative improvements to the product based on end user feedback.

DAC City is a Silicon Valley startup that is developing a machine for capturing carbon dioxide directly from the air for industrial applications. I partnered with the founder to build the prototype for the direct air capture machine. My contribution includes: - Embedded process control software that controls a number of sensors on the machine and processes commands sent by users over Bluetooth and WiFi.- Plans for an application that logs on Co2 captured by machines into an immutable ledger immutable ledger for MRV verification.

GravityAI is a marketplace for data science and machine learning algorithms designed to meet the needs of enterprise companies.I’ve been helping the team at GravityAI build custom data science solutions for their customers. Some of these solutions include:- OCR models for extracting text from checks, purchase orders, invoices, and tax documents.- Tableau dashboard support for a media company- Containerizing popular open source machine learning algorithms to be placed on the marketplace.

Since 2014 theCoderSchool has been teaching kids aged 7-18 how to code though year round after school programs and summer camps. Their secret sauce is an individualized teaching methodology that keeps the students engaged.As a coder coach I provide coaching and project-based instruction to students on a weekly basis a variety of programming languages and Computer Science topics, including Python, Java, JavaScript, C#, SQL, Scratch, and Machine Learning. Sometimes I teach students 1-on-1; Other times I teach in groups of five or more.During the Spring of 2021 I taught a semester of python and scratch programming to a group of 30 gifted elementary and middle school students.Some of the projects I’ve worked on with my students include:- Coding games like Connect Four, Snake, Pong, and Flappy Bird from scratch- Coding the K Nearest Neighbors machine learning algorithm from scratch- Building a Random Wikipedia article generator using GPT-2- A multi-player space battle royal game using Unity- An AWS data pipeline that ingests diluted share data for over 15,000 companies using SEC's EDGAR API.- A rubik's cube solver program

Invenst Labs does Marketing and Operational Innovation for Nonprofit and For-profit companies. Invenst works specifically with companies with strong corporate social responsibility that are out to provide benefits to the environment or to human beings. We are a small interdisciplinary group of people who are committed to changing the world, one bright idea at a time.As a Data Scientist, I gather and analyze data on small businesses, startups, and freelancers in order to identify new business opportunities and also inform the content and marketing strategy for the company. I also develop new software tools and applications that help the company create innovative solutions for clients.

Data Science Consultant at Insight Softmax ConsultingInsight Softmax Consulting provides business-focused solutions using Data and AI. Clients include Autodesk, BMW, BookLawyer, Creative Market, and Zaro Transportation.I developed a machine learning model to predict mobile app user conversions for a South African oil and gas company and to identify characteristics that drive user transactions.I also developed prediction/forecasting models for an international SaaS product.

Traction Tools Software helps organizations implement Entrepreneurial Operating System (EOS) with ease. As a contract Machine Learning Engineer, I used machine learning algorithms to perform a customer churn analysis of the company's customer data to help the data team devise customer retention strategies.

Tesla designs and manufactures electric vehicles, charging stations, and solar and battery products for homes and businesses to accelerate the world's transition to renewable energy.As a member of the Energy Fleet Analytics and Service Engineering Team, I developed data processing techniques to evaluate performance and find failures in solar panels and storage products, wrote programs to automate routine actions and analyses, and developed reports and presentations for upper management, finance partners, and others. Notable accomplishments:- Wrote script to track solar inverter vendor RMA shipments, reducing customer issue resolution time by 7 days and saving Tesla $336,000 a year in lost energy. - Reduced runtime of machine learning algorithms that identify electrical issues in solar systems by 86%.- Collaborated with SMEs to build a data processing pipeline to validate battery cell retention models using historical battery cell signal data. - Used Spark to compute performance metrics over Tesla's 100,000+ battery fleet to drive product improvements where needed.

Fitly is a two-time, digital health award winning company that created SmartPlate, the world's first intelligent plate that instantly identifies and weighs foods using computer vision and deep learning technology. My notable accomplishments include:- Saving over $150K in crowd sourcing costs by creating a machine learning platform that uses deep learning to automatically clean thousands of images for model training. Resulted in a 10% boost in food recognition accuracy of food recognition model.- Helped team build multi-label classification model by creating web applications to annotate millions of food images through the Amazon Mechanical Turk platform.- Increased the size and quality of internal image dataset tenfold by building web crawlers to extract images from multiple sources.- Developed a nearest-neighbor based used behavior model to reduce false positive image classifications.

I worked in a multidisciplinary team of engineers to develop aircraft fire control system software used by warfighters in over 20 countries to identify and track adversaries in the air, on the ground, and in the sea. My notable accomplishments:- Enhanced the portability and testability of the F-35 Targeting System used by 11+ countries onto multiple hardware and software platforms by creating an OS Abstraction Layer.- Eliminated obsolescence issues in integration test assets by upgrading software to display, process, and store streaming data in real time using new hardware components.- Solved critical failures discovered in production by analyzing sensor data.- Helped engineers troubleshoot software failures encountered during flight tests by creating tools to extract and transform aircraft data into formats consumable for analysis.- Conducted $5,000 trade study on facial recognition technology for potential integration into fire control products.

Data Science @ Galvanize's Data Science Immersive ProgramProjects I've worked on include:- Clustering of New York Times articles by latent features using NLP and Non Negative Matrix factorization techniques.- Churn analysis on ride sharing company data in order to build a model to predict customer churn and provide suggestions on improving customer retention.- Hadoop Map/Reduce jobs that uses facebook social graph data to suggest friends to users.- Using Spark to create a Naive Bayes model that predicts category of news articles based on content.- Matrix factorization based recommendation system that recommends movies to users.- An application that uses ARIMA and Holt-winters timeseries models to predict inbound email traffic in a user’s Gmail account.

Helped Harris Corporation evaluate the spatial data processing capabilities of MSSQL, MySQL, and Marklogic by running performance benchmarks on large volumes of data in the cloud.

Helped the Quality Information Systems team build their flagship quality reporting system used by 6,000+ engineers across 15 sites to manage and report quality metrics for over 56 missile and fire control products.

- Automated the installation and configuration of workstations in the Computer Science department using scripts.- Helped System Administrator with basic website maintenance and troubleshooting PC issues.

Created a web service that allows engineers on 40-60 government programs to manage and track action items created during weekly team meetings.

Streamlined the reporting of KPIs presented to senior managers on the High Mobility Artillery Rocket System (HIMARS) program by building a web application that pulls data from various databases.

MotionPoint helps companies scale their businesses to new markets by localizing their websites in different languages and optimizing their performance to engage with those new audiences. The company serves hundreds of brands and organizations, including JetBlue, Humana, ASOS, Verizon, and Amtrak. As a sales engineer, I helped drive the sale of Motionpoint's globalization platform to new customers. I used my web development skills to assist sales managers in generating accurate price quotes on new translation projects based on text data scraped from their client's website.

Enhanced the quality and presentation of the college website across all 12 campuses by:- Fixing HTML errors in webpages- Correcting grammatical errors- Participating in a usability test, which prompted a website redesign to improve user experience- Making sure that webpages conformed to W3C web content accessibility guidelines.- Assisting in the integration of the website into a content management system.