Steven Schmatz

$25M raised to solve global sales tax compliance for accounting firms and finance teams. Built by top industry veterans from Stripe, TaxJar, Fonoa, and more. Lucky to serve and partner with: - Tech and retail companies like Pulley, Omni, Levanta, Italic, Revv, Rosabella, + many more. - Accounting and tax firms like Graphite Financial, Rooled, Rivet Tax, + many more. Backed by XYZ, Headline, Vinyl, Siqi Chen (Runway.com), Mike Rosengarten (OpenGov), and dozens of CFOs/accountants/founders.

We built one of the world's best communities for founders. We raised over $85M from Founders Fund, Tiger Global, and more, and served over 10k founders, changing their lives. Tens of thousands of founders have met their cofounder, raised money, found customers, and more. Our product team was focused on building a community platform curated for founders, driven by data and AI: a founder knowledge base indexing tens of thousands of founder discussions, a matchmaking platform, founder profiles and search, an investor index, our own session library, and more. I joined as one of the first senior engineers reporting to the CTO, became one of the first managers, we grew to ~30 engineers, and later led the product and data engineering functions after the CTO left in early 2022.

(Acq. by Landing) Barsala is a hospitality tech startup for luxury short- and long-term stays and related SaaS tooling. While COVID would make anyone fear running a hospitality company, we pushed through and grew throughout COVID. I joined as a founding engineer and ended up leading our product engineering efforts. We built out a guest-facing mobile app (4.7 stars on App Store!), a web booking platform, a SaaS platform for building owners, as well as various internal tools and systems such as fraud detection.

Worked on Quora Ads Ranking, focusing on improving CTR via deep learning. Developed the feature via matrix factorization methods and integrated it into daily training and serving with Airflow and Spark.

Worked on NLP infrastructure for Search. Built a deep learning model for categorizing web pages. Proposed a new project which was presented at Google's (internal) 2018 Research Conference, and received Google's highest level of peer-awarded bonus for my work.

Trove was a startup building an intelligence layer on top of email to help grow and share your network. Built a business-critical deep learning text classification algorithm to detect introductions over email. 99.0% accuracy and 0.991 F1 score. Convolutional neural network-based approach. Lead entire process from data collection to model deployment. Created a recommendation system to recommend people to meet. Based on a user-level embedding over email messages. Created new method for quality engineers to detect UI components like images in UI screenshots using computer vision methods.

—  Hopelab is a social innovation lab in SF focused on health & wellness of young adults. Hopelab makes a product called Vivibot which is a Messenger bot for young cancer survivors. — Built an analytics dashboard in Tableau for monitoring user growth, retention, and usage of Vivibot. — Built a Python API which processes large amounts of conversation data in parallel and serves it to Tableau. —  Deployed on AWS services.

Working on creating deep reinforcement learning agents to play in Google DeepMind's StarCraft 2 environment, mentored by Prof. Satinder Singh.

— The Tuesday Company provides a software suite for campaigns to utilize social media interaction and data. — Lead design and engineering for a business-critical cross-platform React Native mobile app for volunteers (team of 5 engineers) — Built a React web dashboard for campaigns to view interaction data, trends, and to organize outreach through their volunteers. — Went from product concept to implementation with several political campaigns in three months.

Lead of design and development for a marketplace centered around ethically / sustainably sourced products. Our mission was to provide the best experience possible to shoppers who have the environment and ethics in mind. We aimed to make it easy and appealing to buy one-of-a-kind goods that they could feel good about.

Pluckphoto is an app for sending a print of a random photo from your camera roll to you every month, for free. Lead development or built the majority of: •  The Pluckphoto website •  The frontend iOS app (Swift) •  The stack (AWS)

Bounce Labs is a company which is dedicated to creating mobile apps to affect social change. Our first product, Bounce, is an iOS app to make neighborhoods safer and more social. Like every night at the University of Michigan, you find yourself at the UGLi studying and cramming for that exam, or you find yourself stuck on North Campus after completing another group project. You take a look at your watch and it is 1am. It’s dark. It’s cold and you want to get home safely. What should you do? With our app, you can join and grow communities for your dorm, block, or neighborhood – and bounce home with them later in the night. We started at MHacks V, where we stayed up for two nights in a row designing and building a prototype. Since then, with support from the OptiMize Social Innovation Challenge and the Thiel Foundation, we've been spending almost all of our spare time fleshing out our product. We recently launched on October 16, 2015.

Worked with a team of four to design and build the API powering Weebly’s suite of mobile apps. Wrote proposals, designed API endpoints, implemented them in PHP & JS, and worked with the team to create & use the first integration testing framework for the API. This API is responsible for millions of requests across more than 240k users.

Worked on the landing page and analytics.

Designed and implemented a landing page for a local startup.

Awarded a full $10,000 merit scholarship for summer research at the Research Science Institute, sponsored by the Center for Excellence in Education and held by MIT. Here I conducted an internship at Harvard Medical School's Center for Biomedical Informatics. The project I was involved with studied the relationship between bacterial drug resistance and "protein disorder", or the quality of proteins to have multiple interchangeable tertiary structures. I was involved in the implementation of a neural network-based protein disorder predictor, data wrangling/pipelining, and the statistical methods used to study the correlation.

My work here was focused around the chemical modeling of metalloproteins. We characterized the reduction-oxidation properties of gentisate 1,2-dioxygenase, which is a vital protein to bacterial hydrocarbon metabolism. My responsibilities included running organic synthesis of a biomimetic compound to study its interactions at low temperatures, as well as characterizing the compound with NMR, GCMS, and UV/Vis spectroscopy.