San Francisco State University

Staff Software Engineer at Fivestars

Skills: TypeScript · Node.js · JavaScript · GraphQL · Google Cloud Platform (GCP) · Python · SQL · React.js · PostgreSQL

Neo.Tax

Principal Engineer

Using advanced machine learning Neo.Tax is turning income taxes for businesses into a painless process that gives money back through the R&D Credit.

At Neo.Tax I am managing a team of talented engineers who are building the future of business taxes.  The client uses Typescript, React, AntD, and GraphQL.  The server runs in Google Cloud Platform and uses NodeJS, Typescript, NestJS, GraphQL, TypeORM, Postgres, Redis, Cloud Functions, and MongoDB.

My role includes mentoring and guiding engineers in their career, ensuring projects are delivered on time, working with product to build out the roadmap, plan sprints, establish processes and standards, and coordinate release schedules.  I also still provide architectural guidance, review code, and write code when I am lucky enough to have the time.

Engineering Manager

My role included working with product to build out the roadmap, plan sprints, provide architectural guidance, establish processes and standards, manage releases, review code, and write code.

Tech Lead

At Fivestars I have been building the customer and merchant facing solutions that enable the processing of payments+loyalty and backend services to support these applications.

The customer facing payment solution is a U.I. rich Android application, predominantly written using Typescript, Redux, and Redux-Sagas with 100% test coverage using Jest, along with a bridge to the native layer written in Kotlin.  This application needs to securely process credit card transactions and must be available 24/7 to run a payment.

The merchant facing solution is a U.I. rich embedded WebApp written using Typescript, Redux, and Redux-Sagas with 100% test coverage using Jest.  This WebApp integrates with a multitude of point-of-sale systems running on Windows, Android, and iOS.

The customer and merchant applications communicate over WebSocket, WebRTC, UDP or Bluetooth based on an analysis of the best transport to use.

The backend services are written in Python or NodeJS and are deployed through Kubernetes and Helm.  The Python solutions use a variety of approaches from Django to FastAPI.  The NodeJS stack uses Express, Express-Validation with 100% unit test coverage using Mocha, Chai, Sinon, and Supertest.  The persistent storage is Postgres.  Asynchronous processing facilitated by Kafka.

Fivestars

Staff Software Engineer

Fivestars digitizes loyalty programs for small to medium size businesses and then gives merchants a suite of powerful tools for engaging their customers.  By replacing the punch card with a tablet Fivestars is able to create persistent relationships between merchants and customers.  I am currently developing the next generation version of Fivestars' tablet software using react, redux, and redux-saga.

Senior Software Engineer

WaterSmart provides consumer facing and utility facing websites through a cloud based SaaS platform. Uses behavioral science and algorithms to reduce consumption via a suite of targeted communications across multiple channels.

Engaged in every aspect of the platform as a full-stack individual contributor, from the backend data pipelines and models using Python and Bash through a Pub / Sub framework based on RabbitMQ, to the front-end websites using PHP, Javascript, CSS, and HTML. Data storage is a combination of files, MySQL, and Cassandra. A special focus on complex projects exemplified by the irrigation detection algorithm (patent pending), internationalization of platform to support the United Kingdom, and implementing business critical partnership integrations.

Linux, Python, Bash, RabbitMQ, Chef, Docker, Autoscaling VMs, PHP, JavaScript, CSS, HTML, MySQL, Cassandra

WaterSmart Software

Thomson Reuters is a large Fortune 500 company with a global presence.  Worked in the Financials & Risk division building solutions for companies and individuals who generate income through financial markets.

Successfully managed a group of forty plus Software and QA Engineers.  Located primarily in San Francisco with team members in London, Boston, and New York. Ownership of a number of important analytic content sets including Estimates, Fundamentals, and Pricing.  Structured teams around content sets and effectively introduced Scrum to each team.  Created redundancy and fungibility by cross training across content sets.  Reused key components and frameworks across content sets allowing for rapid delivery.  

Linux, Windows, C++, Java, C#, PHP, Bash, MySQL, SQL Server, Oracle RDBMS, HBase, Web Services, Pub / Sub

Thomson Reuters

Head of Analytics Development

Initial team of four developers which expanded over time to support increase in projects. Included building out the next generation Fundamentals data system to replace two legacy systems. A multi-year endeavor, involved working with teams from New York, London, Minsk, Bangalore, Beijing, and Manilla. Became an expert working effectively with global teams. Built a calculator service that could take an arbitrary collection of equations, create a graph of equations, and calculate each equation by replacing variables with input data. 

Windows, C++, Boost, C#, SQL Server, Socket, Multithread, Jenkins

Software Development Manager

StarMine was acquired by Thomson Reuters in 2008. Integrated the StarMine stack into the Thomson Reuters' data centers. Initially built a data transfer solution to move many terabytes of data from San Francisco to the East Coast without any down time. Built replication systems to support immediate failover. Determined how to scale systems to work at the enterprise level. Introduced many of the ideas from StarMine into the larger Thomson Reuters ecosystem.

Linux, C++, PHP, Bash, rsync, MySQL, Oracle RDBMS, Pub / Sub

Sr. Technical Specialist

StarMine provided ratings of broker analysts as well as hedge fund style quantitative models to fundamental managers.  The company started by overweighting accurate analysts to create a SmartEstimate for future earnings.

Worked on every aspect of the platform from backend data processing to front end presentation on the website. Productized many quantitative models, introduced a text search engine for all analyst reports, implemented algorithms to determine the accuracy of analysts, and scaled the backend data pipeline.

Linux, C++, PHP, Bash, MySQL, Pub / Sub, JavaScript, CSS, HTML

StarMine Corporation

Vividence provided online focus groups that visited client websites and through a browser plugin were given objectives while quantitative and qualitative information was captured.

Decoupled infrastructure through Pub / Sub using MSMQ, asynchronous processing allowed high load to be handled gracefully. Worked on free text sentiment clustering algorithm and analyzing / visualizing click stream data.

Windows, C++, Visual Basic, SQL Server, MSMQ, ASP with JScript, XML, XSLT, SOAP, JavaScript, CSS, HTML

Gareth Ivatt