Head of Data Science

• Feb 2023 - Present

Building software that will change the landscape of portfolio management from the ground up. Our website: https://avayl.tech/ Building software that will change the landscape of portfolio management from the ground up. Our website: https://avayl.tech/

Founder in Residence - BE10

• Sep 2022 - Jan 2023

Entrepreneur First (EF) is a startup incubator/accelerator with +10B in portfolio value, +3000 alumni and +500 companies founded. I was 1/50 selected from +1000 applicants. Entrepreneur First (EF) is a startup incubator/accelerator with +10B in portfolio value, +3000 alumni and +500 companies founded. I was 1/50 selected from +1000 applicants.

Researcher

• Mar 2022 - Jan 2023

Working on various research projects both internal and with partnering organizations. Project Autonuoma: Creating and deploying Machine Learning models to predict the amount of ambulance calls per district across each city in the country. Delivering a fast, expandable, and accurate ML solution with an extensive research report. Achieved prediction accuracy within 0.27 calls/day. Project Demino: An app for dementia patients. POC phase. Usage of large amounts of structured and unstructured data related to dementia patients for a viable and helpful solution with minimal supervisor intervention. Performed research on possible solutions and application of large amounts of data collected from dementia patients. Show less

Data Scientist

• May 2022 - Sep 2022

Improving performance and metrics of the current ML production models. Evaluating different model performance and metrics and presenting them in a clear and concise way. Working in SCRUM/AGILE development environment to create easily integratable AI solutions. Performing model hyperparameter optimization using AutoML and custom parameter sweeping. Performing data preparation tasks including labeling, data extraction, data preprocessing, etc. Refactoring POC scripts into concise, DRY, performant, functional solutions. Developing POC solutions for various low and high level problems. Participating in code reviews and leaving code improvement suggestions. Presenting a company wide tech talk about the declarative programming paradigm. Show less

Research Assistant

• Nov 2021 - Jun 2022

Implementing biologically inspired neural network models into The Virtual Brain (TVB) architecture to model large scale parkinsonian brains to be used in clinical studies.Working on large scale radiomics, morphometric and psychological data to predict psychiatric and motor outcomes in Parkinson's disease patients undergoing deep brain stimulation (DBS) procedure. Finding relationships between different groups of data. Presenting results in a clear, concise and understandable manner.Working closely with professionals from different fields (neurosurgeons, neurologists, psychologists, physicists) to suggest further research topics and areas.Working on anonymizing and sharing MRI and CT scan data with partners abroad. Constructing structural large scale brain connectomes from MRI and CT scan data to be used in large scale brain simulations.Dopaminergic neural mass model curation and usage in large scale brain modeling. Model fitting and validation using EEG data features and ML models.Working on large scale brain simulations to improve the placement and stimulation parameters of DBS electrodes in patients with PD. Show less

Research Intern

• Sep 2020 - Nov 2021

Freelance Software Engineer

• Sep 2015 - Nov 2021

Working on various projects of differing sizes and scopes with multiple technologies and stacks. Writing performant, DRY, compositional, functional code for both backend, frontend and full-stack solutions. Some of the projects I worked on: Project iTrade.gg: A website for trading Rust skins. I was responsible for the backend and minor front end changes. Using the data from historical item prices to fight against instability, arbitrage, etc. Significantly improved back end performance, security and added a substantial amount of new functionality in terms of marketplace item price and availability control. Implemented a self sustained and improvable ML solution. Python was used as the backend solution for the stack. Project ArmyFight: A web app for large historical battle simulations using native JS, CSS and HTML5 working jointly with the project owner. I worked on backend optimization using native JavaScript improving simulation capabilities from 10's to 100's and 1000's of entities. Also worked on minor front end changes and additions. Show less