Evan Radkoff, Developer in Seattle, WA, United States
Evan is available for hire
Hire Evan

Evan Radkoff

Verified Expert  in Engineering

Machine Learning Engineer and Developer

Seattle, WA, United States
Toptal Member Since
January 29, 2021

Evan is a former Amazon engineer specializing in applying machine learning and data science to solve the most difficult problems. He is the rare full-stack data scientist who can deliver on the entire ML engineer lifecycle from exploratory data analysis to model tuning to deployment and operations. Backed by an MS in CS, Evan also ramps quickly on new projects and deals well with ambiguity. Evan loves working with natural language data and is passionate about the intersection of AI and music.


Self-employed, Sounds and Words LLC
Pandas, Scikit-learn, PyTorch, App Development, Swift...
Sidetrack AI
Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP)...
Amazon Web Services (AWS), SQL, HTML, Spring, Hibernate, JavaScript, CSS, Scala...




Preferred Environment

Amazon Web Services (AWS), Scikit-learn, PyTorch, Slack, MacOS, Jupyter Notebook, Python, PyCharm

The most amazing...

...thing I've developed is a GAN-based generative model that outputs drum patterns that sound good enough to be used as percussion in actual songs.

Work Experience


2020 - PRESENT
Self-employed, Sounds and Words LLC
  • Offered contracting and consulting services for gaming, finance, data recovery, and defense businesses.
  • Released an app in the macOS App Store that intelligently splits album audio files into tracks, reaching #72 on the music app charts.
  • Developed classification models for drum sounds, using deep learning, for a future product.
  • Trained a GAN-based generative model that writes original drum patterns for use in electronically produced music.
Technologies: Pandas, Scikit-learn, PyTorch, App Development, Swift, Artificial Intelligence (AI), Python, Data Science, Time Series Analysis, Data Mining, Data Modeling

Machine Learning Engineer (Contractor)

2019 - 2020
Sidetrack AI
  • Wrote a Python module that ingests email data and intelligently segments emails into smaller parts using a CRF and LSTM.
  • Discovered an optimal solution for clustering business documents based on multimodal data. Established evaluation metrics to ensure quality clusters.
  • Modeled high-level features of business documents with text classification methods.
  • Researched and implemented several state-of-the-art keyphrase extraction methods, including sequence-to-sequence generative neural networks, adapting them to a new domain.
  • Adapted an open-source data annotation software to fit our needs. Oversaw the collection of thousands of labels for multiple ML tasks.
Technologies: Generative Pre-trained Transformers (GPT), Natural Language Processing (NLP), GPT, Research, Pandas, Graphs, Neural Networks, SciPy, NumPy, PyTorch, Scikit-learn, Artificial Intelligence (AI), Python, Data Science, Data Modeling

Software Development Engineer

2014 - 2018
  • Created a back-end service for authors to manage events related to marketing their books.
  • Designed and implemented back-end APIs for a data dashboard that lets authors see how their books are selling and how readers are engaging. Data is pulled from several marketplaces worldwide and aggregated to multiple time periods.
  • Delivered a core feature to Author Central, a platform for authors to manage their presence on Amazon that allows authors to associate books with their profile.
  • Responded to several service outages by diagnosing systems in real-time and providing a speedy path to recovery. Wrote post-mortems of outage events and put fixes in place to prevent recurrences.
  • Helped build and maintain deployment pipelines for over a dozen services.
  • Launched Write On by Kindle, a story lab and community for writers and readers.
  • Secured user data by integrating with Amazon-internal encryption solutions.
  • Saved the company tens of thousands in monthly costs by initiating an audit of servers and ETL job use.
  • Gave a lecture internally about the Scala Collections API and functional programming.
Technologies: Amazon Web Services (AWS), SQL, HTML, Spring, Hibernate, JavaScript, CSS, Scala, Java, Amazon DynamoDB, NoSQL, Distributed Systems, Servers, API Design

Drum Sound Classification

A Python module presenting my research utilizing machine learning to classify 1-shot drum sounds.

Modern music production can involve incorporating drum one-shots from large, disorganized libraries of sounds. I am working on AI methods to help effectively navigate such libraries, which includes dealing with mislabeled or inconsistently labeled files. Not only do neural networks solve this problem, but I've found that the intermediate embeddings they produce are useful for other related problems. This work is part of the research and development for a future product, released in the meantime as a blog post and open-source project.

Album Split macOS App

A macOS app that intelligently splits album audio files into individual track files through a simple drag-n-drop interface. This was built for music archivists who are digitizing albums and do not wish to split recordings into track files manually. The app reached #72 in the App Store in the music category and a pro version is in development.
2012 - 2014

Master's Degree in Computer Science

University of Wisconsin-Madison - Madison, WI, United States

2008 - 2012

Bachelor's Degree in Mathematics and Computer Science

The College of Wooster - Wooster, OH, United States


Scikit-learn, PyTorch, NumPy, Pandas, TensorFlow, SpaCy, Matplotlib, SciPy, Keras, NetworkX


PyCharm, Git, Slack, Google Compute Engine (GCE)


Python, Java, Scala, SQL, CSS, JavaScript, HTML, Swift, Bash Script


Data Science, App Development


Hibernate, Spring


Jupyter Notebook, MacOS, Amazon Web Services (AWS), Amazon EC2, Linux, Docker, Google Cloud Platform (GCP)


Databases, Amazon DynamoDB, Amazon S3 (AWS S3), NoSQL, MySQL, PostgreSQL


Clustering Algorithms, Artificial Intelligence (AI), Machine Learning, Natural Language Processing (NLP), Music Information Retrieval (MIR), Convolutional Neural Networks (CNN), Text Classification, GPT, Generative Pre-trained Transformers (GPT), Algorithms, API Design, Distributed Systems, Neural Networks, Graphs, Research, Regular Expressions, Audio, Librosa, Deep Learning, Automated Summarization, Keyphrase Extraction, Data Mining, Deep Neural Networks, APIs, Time Series Analysis, Data Modeling, Computer Vision, Compilers, Servers, Digital Signal Processing, Custom BERT, Generative Adversarial Networks (GANs), Variational Autoencoders, CI/CD Pipelines, Google BigQuery, Reinforcement Learning, Deep Reinforcement Learning

Collaboration That Works

How to Work with Toptal

Toptal matches you directly with global industry experts from our network in hours—not weeks or months.


Share your needs

Discuss your requirements and refine your scope in a call with a Toptal domain expert.

Choose your talent

Get a short list of expertly matched talent within 24 hours to review, interview, and choose from.

Start your risk-free talent trial

Work with your chosen talent on a trial basis for up to two weeks. Pay only if you decide to hire them.

Top talent is in high demand.

Start hiring