Brown University

Harvard University

South Shore Charter Public School

Quantic School of Business and Technology

Research Data Scientist at Deepcell, Inc.

San Francisco Bay Area

Machine Learning consultancy helping companies, organizations, and individuals build, improve, and deploy their machine learning applications.
-  Managed ML research and deployment strategies for 100+ offerings in B2B and B2C.
-  Built 100+ ML pipeline analytic processes to decrease clients’ costs by 20-45%.
-  Worked with customers in Music, AR, VR, including a CES Product Launch that generated 10k users.

5cubeLabs

Co-Founder, CTO

Co-Author on "Developing Trustworthy Machine Learning" (early access)

O'Reilly Media

Book Author

On Deck Longevity Biotech (ODLB) is a continuous community for people to come together to build, join, or invest in revolutionary longevity biotechnology startups.

On Deck

ODLB, Longevity Biotech Fellow

Paying it forward after success with Bitcoin, Ethereum, & Stellar. Angel Rounds in various early-stage stealth biotech and AI companies.

Stealth

Angel Investor

Selected as one of 100 participants out of thousands of applications for a top tech talent fellowship. Alums include former C-suite execs, directors and VPs at Khan Academy, Flexport, Google, Crunchbase, Lyft, Uber, Postmates, Foursquare, and Eventbrite.

ODF4, Fellow

Program set up to pair talented senior engineers together as company founders. Fellows were granted housing in NYC, advisors, along with an initial $25,000 investment (at $1,000,000 valuation) in their companies by Backend Capital (<a href="https://www.crunchbase.com/organization/backend-capital">https://www.crunchbase.com/organization/backend-capital</a>)

Major League Hacking

Hacker Fellow

Multi-disciplinary team of scientists and engineers. Collaborators include researchers from research institutions such as Google Brain, University of Oxford, and Vector Institute for Artificial Intelligence. Our published experiments and tools can be found on <a href="https://github.com/for-ai">https://github.com/for-ai</a> .

FOR.ai

Machine Learning Researcher

UnifyID is building a revolutionary identity platform based on implicit passwordless authentication. Their solutions allow people to identify themselves in ways that are extremely difficult to forge or crack (even with information from a source like the Equifax hack).
- Robust neural architecture optimization infrastructure
- Identification algorithms for extremely-unevenly-sampled data scalable to millions.
- Bleeding-edge ML research at the intersection of signal processing and deep learning.

UnifyID

Machine Learning Engineer

- Contributed to Bayesian Analysis & Probabilistic programming toolkits for within Tensorflow.
- Reconstructed machine learning papers with, wrote documentation for, and wrote GPU/TPU-parallelizable examples for Tensorflow Probability applicable to quantitative finance, astrophysics, and neurology.

Google

Software Engineer

- Constructed IPFS-based content-delivery network (using React.js, Redux, Electron.js, Go, the Tor network, 1m5) for censorship-resistant journalism for cross-platform use by news reporters in countries and regions with high levels of government surveillance (e.g., Syria, Iran)

- Developed proofs of concept for the decentralized search of IPFS nodes (using ipfs-tika, ElasticSearch 5, RabbitMQ, Go, and Node.js) and client-side NLP tools (using PyTorch, Gensim, & NLTK) to safeguard against troll-farms, DDoS attacks, and state-funded bad actors.

Inkrypt

Software Engineer (IPFS, Mobile)

- Developed distributed facial detection system for improving real-time facial recognition accuracy among 50-enterprise-customer-bandwidth (with ~200 security cameras each) using Apache Spark, Apache Kafka, OpenCV, Hadoop, Amazon EC2, and TensorFlow. System allowed for testing edge cases of algorithms developed in academia in real-world environments (such as against atmospheric or lighting artifacts).

- Prototyped algorithms for reducing racial bias in applications of facial recognition to security footage and for making outputs explainable for the purpose of diagnosing adversarial attacks.

Suspect Technologies

Machine Learning Engineer (Facial Recognition)

- Improved multi-audio-feature-detection algorithms for use in harmonic mixing software (used by top musicians like Steve Aoki) from 90% to 93% on validation sets by using hybrid models and using Bayesian optimization to save time on hyperparameter tuning.
- Modified trained TensorFlow models to be compatible with legacy systems written in C# (Microsoft .NET Framework).

Mixed In Key

Deep Learning Engineer

- Developed bleeding-edge federated machine learning & homomorphic encryption tools that both allowed for sophisticated machine learning insights, while also protecting user data from the hacking or 3rd-party abuse that plagues mainstream social networks.

-  Led creation of econometric and app-usage data analysis strategies for private-betas, A/B testing, and for improving user churn-rate.

-  Set up Stellar (XLM) testnets and protocols for verifying millisecond-scale Lightning protocol sidechain transactions on React Native mobile apps (upwards of 128x faster than Ethereum verification at the time).

Hellofriend

Data Scientist

Advised Team on best strategies for non-gimmick machine learning application, user acquisition, and analytics. App was bought by Hubspot.

Evolve App (Acquired by HubSpot)

Advisor

- Used Python, LabVIEW, and mammalian cell culture skills to enhance laboratory techniques involving two-photon microscopy of prefrontal cortex primary neuron cultures at MassGeneral Institute for Neurodegenerative Disease (MIND).

- Used knowledge of Alzheimer's Disease, AAV vectors, and mouse surgery to investigate the role of calcium signaling in astrocytes in the progression of neurodegenerative disease.

- Applied understanding of apoptosis, smooth muscle atrophy, RNA laser-capture, and RNA sequencing to design project for investigating transcriptomic patterns in blood vessel linings of brains with cerebral amyloid angiopathy.

- Used understanding of OpenCV, tight junctions, cryosectioning to investigate the ability of PEPDART molecules to permeabilize the blood-brain barrier as a means of allowing other small molecule therapeutics access to the central nervous system.

Massachusetts General Hospital

Multi-photon Microscopy Research Technician

- Used knowledge of genetic algorithms, Python, and stem cell biology to create a simulation of industrial-scale extracellular vesicle production that could simulate continuous exponential expansion instead of discrete stepwise expansion.

- Used R and Javascript to produce figures representing data, as well as contribute to the manuscript as a co-first author (more information coming soon, currently in review by ACS Nano).

Massachusetts Institute of Technology

Karp Lab Stem Cell EV Researcher

- Used genetic algorithms and machine learning models of co-factor pathways alongside directed evolution technologies to create novel and viable bacterial strains for secreting molecules en masse (for 100X reduction in the cost of naturally-occurring molecules like nicotinamide mononucleotide).

- Company spun-out from project aimed at creating cheaper radioresistance drugs for use in cancer patients undergoing radiation therapy. Drugs and compounds were identified through research at Harvard Medical School and Brown University.

- Received financial support from MIT Sandbox and private angel investors, as well as equipment and laboratory facilities support from Boston University, IndieBio, BosLab, and Brigham & Women's Hospital.

Project AgeTuneUp

Co-Founder & CTO

This position in the Rand Lab involved investigating the combined effects of mitochondrial genotypes, nuclear genotypes, and diet on aging in D. melanogaster in experiments funded by the Howard Hughes Medical Institute (HHMI).

- Ran starvation-based demography assays and climbing assays to assess health-span of subjects, a system designed to require less time for testing anti-aging interventions than traditional survivorship assays.
- Did triacylglyceride, glucose, and mitochondrial oxygen consumption assays to test validity of procedure.
- Developed new mathematical models for healthspan and collecting statistics on the parameters, eliminating problem of using models that didn’t account for rate of health deterioration.
- Used knowledge of databases, genomics, and networking to assemble compendium of anti-aging interventions across multiple phylogenies for predicting possible combinatorial therapies for thesis project.

Rand Laboratory Research Assistant, HHMI Scholar

- Co-founded synbio startup for creating strains of Lactobacillus acidophilus that could be used for multiple-toxin-detection in dairy products. The goal was a cheaper method of food inspection for use in countries with less funded or developed FDAs. This was done by leveraging resources and techniques of MIT and networking with the Cambridge area DIY bioengineering community.

-  Created multi-logic-gated design for massively reducing the costs of food-safety testing for governments and private companies.

-  Created multiple working prototype strains of the bacterium & Raised funds through a combination of crowdfunding and proceeds from selling affordable custom lab equipment. Received letter of intent from the Thailand FDA.

McAteer LifeLabs

Co-Founder, COO, Researcher

Cryptocurrecy brain-wallets, made simple enough for your regular wallet.

ID Bunker

CTO, Co-Founder

- Took part in the creation of the caDNAno software; a program designed to calculate self-organizing DNA sequences for constructing nanoscopic shapes designed by the user.
-  These sequences would attach to a larger template sequence and then self-assemble by means of interactions between base pairs. This was based on DNA origami research that had taken place previously at the Wyss Institute.
-  Learned how to assess successful assembly of 2D and 3D DNA nanorobots with macromolecule separation analysis.

Wyss Institute for Biologically Inspired Engineering

Research Intern, Programmer

- Constructed interactive online experiments (iLab initiative) for MIT's Open CourseWare. These include the NXT Program Uploader iLab and the HELEx fuel cell iLab.
-  In a computer science lab typically occupied by graduate students, post-docs, professors, and occasionally undergraduates, I was the first high school student to work as a full-time paid intern in the center.

iLab Experiment Programmer, Robotics Intern

- Created a finite element test based on a model of a human knee to predict which positions would make an anterior cruciate ligament injury more likely.
-  This was constructed from MRI images of the human knee joint.

Northeastern University

Finite Element Simulation developer

- Researched methods of biosequestration for creating an micro-algae bloom in the Southern Hemisphere with Iron (III) Oxide and Synechococcus algae.
-  This was funded by a grant issued by the Technology Academy of Finland.

University of Helsinki

Carbon Sequestration Researcher

- Analyzed plankton composition of waters around Buzzard's Bay, MA.
-  Produced data from microscopy detailing plankton density and diversity in three areas with distinct levels of boat traffic.

Woods Hole Oceanographic Institution

Research Volunteer

- Youngest person ever to pass center’s veterinary student internship exam (Age 14). Matriculated from volunteer position.
-  Procuring correct medications, preparing microscope slides, and running tests on fluid and tissue samples for mammalian, avian, reptilian, and amphibian patients while acting as lab assistant for resident Veterinarian.
-  Assisted with multiple surgeries (highlights include re-suturing in R. norvegicus and an ovarian cyst removal from C. porcellus) by monitoring vital signs, procuring necessary medications, maintaining the sterile field, and maintaining anesthesia composition.

Matthew McAteer

Matthew McAteer