Paul Filitchkin

Sr Data Scientist

Creative Market

April 2016 - Present

Creative Market is a dual-sided marketplace for design where independent creatives from around the world sell their fonts, graphics, templates, 3D models and more. It's like the Etsy of digital assets.

At Creative Market I use search, recommendation, and machine learning tech to help customers discover products while empowering independent creators to succeed.

In July 2017 Creative Market spun out of Autodesk and became an independent company with a $7M Series A.

---

Responsibilities

• Maintain core full-text search library and run A/B tests on search ranking improvements
• Develop collaborative filtering recommendation engine based on recent research
• Provide business and user analysis using Python notebooks to help leadership team make decisions
• Build machine learning prototypes using Keras/Tensorflow to prove out new technologies

Accomplishments

• Increased search purchase rate by +35% through multiple A/B tests over the course of a year
• Ran personalized recommendation A/B tests which increased email click through rates by +65% and purchases by +39%
• Supported SEO growth (single biggest channel at Creative Market) with the help of machine learning
• Developed multiple production ETL pipelines (operating on over a billion records stored in Redshift)

Tech

• Python - NumPy, Scipy, TensorFlow, Keras, Gensim, Jupyter Notebooks
• PHP - Laravel, Guzzle, PHPUnit
• AWS - Elasticache, S3, Redshift, RDS
• C++ - OpenCV, Boost
• Elasticsearch, MySQL

Links

Search Engineer / Sr Web Engineer

Creative Market / Autodesk

April 2016 - July 2017

I joined Creative Market when it was part of Autodesk because of the amazing team, the rewarding mission, and the team's ability to make independent decisions inside of a large company. My initial role was to own the backend search experience and support key SEO projects under the guidance of top SEO consultants.

During this time I independently applied modern machine learning for product discovery, and this became the foundation for the discovery team.

In addition, I made key contributions to custom on-the-fly image resizing microservice based on OpenCV and open sourced the supporting Go library (links above).

CTO

Snapwire

October 2012 - March 2016

---

Snapwire is a platform for on-demand request-based photography assignments where someone can write a brief, name a price, and let its community of photographers submit their best photos. In addition, Snapwire hosts a large catalog of searchable royalty-free photos.

I was part of the founding team that took Snapwire from an idea to a fully fledged dual-sided marketplace for hundreds of thousands of users. Since inception I worked as part of a small team to architect and develop the platform to support the website, iOS, and Android app.

As the company grew I took on the responsibility of managing 4 engineers using agile scrum methodology.

This role also granted me extensive experience with a range of AWS cloud services include: EC2, S3, RDS, CodeDeploy, Route 53, SNS, and Cloudwatch.

During this time I had the opportunity to meet some of our users face-to-face through a series of Snapmeets. I asked about their pain points and got feedback about the app, and to this day it has been one of the most humbling and rewarding experiences of my career.

Responsibilities

• Design and develop core backed libraries (with strong focus on full-text search)
• Design and develop mobile API
• Develop initial iOS app and manage app store listing
• Setup, maintain, and monitor AWS infrastructure

Accomplishments

• Snapwire iOS app was featured worldwide in the photo & video category
• Scaled platform from 1 to 250k users

Links

Analyst

Toyon Research

January 2010 - September 2012

I started working at Toyon while completing my masters degree, and the experience introduced me to a wide range of challenges involving computer vision and embedded systems. Toyon operates largely on small business innovative research (SBIR) grants which aim to push the boundaries of applied research.

During this time I worked primarily with highly trained (PhD & MS) engineers on many different research oriented grants that each posed unique constraints.

Responsibilities

• Maintain telemetry and control server on UAV guidance system
• Develop telemetry analysis and log playback GUI application
• Develop OpenGL accelerated vision tracking solution for ARM processor (there were no embedded systems with OpenCL or CUDA at the time)
• Attend UAV flight testing trips (required fast paced development in harsh environments)

MS, Electrical and Computer Engineering

University of California Santa Barbara

2009 - 2011

Focus in control systems and signal processing. Wrote thesis on autonomous quadruped navigation using computer vision.

BS, Electrical and Computer Engineering

Oregon State University

2004 - 2009

Pursued a self-designed electrical and computer engineering major with a focus in robotics/artificial intelligence and a minor in computer science. Finished coursework in 4 years and did 2 six month internships as part of a university co-op program (MECOP).