Ryan Walker

Most growing companies are either overpaying for outside counsel or avoiding legal work altogether because it's too slow and too expensive. General Legal exists for both of those companies. Experienced attorneys, AI-powered speed, flat fees. If you need outside legal counsel that actually moves at the speed of your business, reach out.

Casetext helps millions of people every month access high-quality, legal research at an affordable price. By combing cutting edge machine learning and search technology with a beautiful and easy to use interface, our platform is disrupting the traditional legal research industry.

I lead the data science and engineering team at Casetext. We're a multidisciplinary, product-driven team building the data backend for Casetext's groundbreaking legal research platform. We develop legal-tailored ML solutions for search, topic modelling, tagging, and recommendation. We build legal data processing pipelines and software solutions to extract valuable new insights from the entire corpus of US law.

Architect and principal developer of CARA, Casetext's flagship recommendation engine. CARA helps lawyers to write and check legal documents by recommending relevant case law. Other responsibilities: - Scalable, real-time recommendation services in Scala and Rust - Document parsing, NER, topic modelling, and classification with Apache Spark - Core NLP development (Scala, Java) for named entity recognition - Search engineering with Elasticsearch and raw Lucene

Designed and implemented Straw, a scalable platform for real-time search on streaming data. Straw allows users to register complex, Lucene-style queries to follow topics and ideas in real-time. As the documents (tweets, new articles, etc.) enter the stream, Straw delivers matching documents to subscribed users with millisecond latencies. The platform is implemented in Java and Python and uses Lucene-Luwak for highly optimized streaming search. Kafka handles the real-time ingestion and the search work is distributed using Apache Storm.

Primary server engineer for Calio, a health and wellness personal assistant built on a custom natural language processing (NLP) platform. - Designed and implemented major server-side features for Calio using NLP technology in Lua, C++, MySQL, and Ruby. - Added features and optimizations to Fluential's core NLP C++ dialogue engine - Designed and trained SVM and other models for building product taxonomies and improving semantic concept recognition in Python and C++. Worked with NLP experts on feature extraction, text normalization, SVM training, tagging, search, and language modelling - Working with CTO and Chief Architect, designed the architecture for a REST API to expose Fluential’s NLP tools and wellness tracking technologies. Led the technical implementation of the REST API - Designed Calio’s Elasticsearch index and its AWS deployment - Built a prototype for a native iOS Calio with an embedded C++ NLP engine and dynamic Lua interpreter. Dynamic code execution on the iPhone probably couldn't pass through an App Store review but, it was an awesome learning experience.

Designed, implemented, and supported a market sizing and forecasting platform to model the California health insurance market during the role out of Obamacare. This platform used non-linear optimization algorithms to combine data from 17 different market sources and produced on-demand executive reports at a highly-refined geographic/product segment scale. The platform reshaped executive strategy around market share and the model forecasts proved highly accurate. Other responsibilities: - Data engineering, feature extraction, and model building for targeted messaging campaigns. Trained and tuned logistic regression models using large, individual-level health and behavioral data sets. Helped team obtain significant lift in conversion with a flat marketing budget - Statistical testing, consulting, and analysis in SAS and R. Complex SQL development to produce custom reports on Teradata and Oracle platforms. Helped design and build team's marketing data warehouse and created many executive-level dashboards with Tableau

- Research in 1D inverse scattering, inverse problems, and mathematical tomography - Graduate coursework and projects in a range of theoretical and applied mathematics topics, including numerical analysis, linear algebra, differential equations, mathematical tomography, and computational fluid dynamics. - Award-winning primary instructor for many college mathematics courses including multivariable calculus and several semesters of ordinary differential equations.

Developed a massively parallel implementation of an image reconstruction algorithm for an experimental tomographic imaging system in the laboratory of Prof. Sarah Patch.

Dissertation: "On A Paley-Wiener Theorem for the ZS-AKNS Scattering Transform"

Masters exam on geometric measure theory

Undergraduate Thesis: "Does Foreign Direct Investment Cause Convergence? Evidence from Panel Data Analysis"