Machine Learning Developer in Vancouver, BC, Canada

Exposes an API for streaming real-time market data through ZeroMQ's pub/sub messaging model. This authenticates using a session based middleware.

Technologies: Node.js, Express.js, Java 8, Ratpack, CDI (Guice), Jackson, ZeroMQ, Memcached, Git, GitHub, Travis CI

Handled a REST API for accounts, users, and trading. The project acts as a lightweight proxy and includes some lightweight logic around parameter validation, permissions, etc. It has integration with an internal SOAP API running on a JBoss applications server.

Technologies: Java 8, Java EE (JPA/Hibernate, CDI, JAX-RS, JAX-WS), Dropwizard, MySQL, JUnit, Mockito, Git, GitHub, Travis CI, Sonar, PMD, FindBugs

Delivers market data. Endpoints consist of real-time and delayed market data, history, time-sales, and streaming session endpoints. It has integration with a third-party SOAP API from where the market data is retrieved.

Technologies: Java 8, Java EE (JPA/Hibernate, CDI, JAX-RS), Dropwizard, MongoDB, Memcached, DxFeed SDK, JUnit, Mockito, Git, GitHub, Travis CI, Sonar, PMD, FindBugs

Exposes an admin console for market data entitlements and reporting

Technologies: JavaScript (ES6), Ember.js, Node.js, Bootstrap, Bower, Git, GitHub, Travis CI

Allows/restricts delayed or real-time access to products/data feeds for both the regular and the streaming API.

Technologies: Java 8, Java EE (JPA/Hibernate, CDI, JAX-RS), Dropwizard, MySQL, Redis, Akka, JUnit, Mockito, Git, GitHub, Travis CI, Sonar, PMD, FindBugs

Front-end layer that communicates with several APIs allowing advisor users to perform financial operations.

Technologies: JavaScript (ES6), Ember.js, Node.js, Bootstrap, Bower, Git, GitHub, Travis CI

Made accounts and orders and the admin services necessary for those processes.

Technologies: Java 8, Java EE (JPA/Hibernate, CDI, JAX-RS), Dropwizard, MySQL, Retrofit, Java Mail, Quartz, AWS SDK, JUnit, Mockito, Git, GitHub, Travis CI, Sonar, PMD, FindBugs

Opens and funds accounts and the admin services necessary for those processes. Most operations require integration with third-party SOAP/REST APIs.

Technologies: Java 8, Java EE (JPA/Hibernate, CDI, JAX-RS), Dropwizard, MySQL, Retrofit, Java Mail, Quartz, JUnit, Mockito, Git, GitHub, Travis CI, Sonar, PMD, FindBugs

An open source project, still under development, to help language learners by mixing subtitles in two different languages. It's possible to choose a percentage of phrases of the target language the user wants to mix together with its native language.

Technologies: MEAN stack, NPM, Bower, OpenSubtitles API

An SLA system used by large call center companies. It became a product so it is possible to run on Windows or Linux servers. As for databases, it's possible to use Oracle, PostgreSQL or SQL Server. It has a High Availability mode using JBoss AS Clustering and Load Balancing through Apache(mod_jk). Its UI was improved using Bootstrap 3.

Technologies: Java EE 6 (EJB, Hibernate, JAX-WS, JAX-RS (RESTEasy), CDI, JSF (RichFaces)), JUnit, JBoss Seam, Maven, Oracle, SQL Server, PostgreSQL, Bootstrap 3, Google Charts

A system to manage the execution of services exposed on a Websphere ESB in order to load data. The project was developed for the largest health insurance company in Brazil, Unimed.

Technologies: AngularJS, Grunt, Karma, NPM, Bower, E2E tests, HTML5, Booststrap, Java EE (EJB, JAX-RS (Jersey), CDI), JUnit, Maven, JBoss AS

A project developed as part of my master's thesis. It's an evolutionary algorithm (EA) called Grammatical Evolution (GE) that has inspiration on the creation of proteins as well as the natural selection. EA is a subfield of machine learning/computational intelligence. This algorithm was used to optimize the prediction of software effort estimation. The GE algorithm generates executable programs in any language, in this case, in Python. It used a well-known dataset called ISBSG to train and validate the model. Before running the algorithm, this dataset was cleaned by removing outliers and other statistical tests. This process was done using R.

Technologies: Python 2.7 (Pandas, Numpy, Matplotlib), Spyder, R

A system to calculate the commission of authorized partners, individual sellers, and teams of the largest health insurance company in Brazil, Unimed.

Technologies: AngularJS, Grunt, Karma, NPM, Bower, E2E tests, HTML5, Booststrap, Java EE (JPA, EJB, JAX-RS (Jersey), CDI), JUnit, Maven, JBoss AS, Oracle, Oracle PL/SQL

A system to assist doctors and patients in order to anticipate possible diseases. This project was developed for the largest health insurance company in Brazil, Unimed.

Technologies: AngularJS, Grunt, Karma, NPM, Bower, E2E tests, HTML5, Booststrap, Java EE (JPA, EJB, JAX-RS (Jersey), CDI), JUnit, Maven, JBoss AS, Oracle, Oracle PL/SQL

A system to create dynamic forms used by doctors in order to measure the risk level of a patient. This project is used by many systems though RESTful API as well as AngularJS directives. It was developed for the largest health insurance company in Brazil, Unimed.

Technologies: AngularJS, Grunt, Karma, NPM, Bower, E2E tests, HTML5, Booststrap, Java EE (JPA, EJB, JAX-RS (Jersey), CDI), JUnit, Maven, JBoss AS, Oracle, Oracle PL/SQL

A system to store and manage all the bidding announcements on the state of Minas Gerais, Brazil.

Technologies: Java EE (JPA, EJB, JSF (PrimeFaces), JAX-WS), Spring Security, Maven, JUnit, Oracle, JBoss AS.

An innovative project to accelerate projects on the early stages by creating a reusable and flexible UI library based on Bootstrap.

Technologies: Bootstrap, SASS, HTML5, CSS3, JavaScript/jQuery

An innovative project recognized by the company as the best idea of 2012. This project attracted government funding to be expanded and to update its initial architecture. Summarizing, RAST is a requirements traceability tool that binds source code and business requirements so that it's possible to easily keep track of the development phase. It provides the completeness of use cases or user stories, impact analysis, automatic traceability matrix, project evolution, and more. RAST is able to trace business and source code written in any language.

Technologies: Java EE (CDI, JSF (PrimeFaces)), Google Guava, JavaFX version

A city hall project that allows citizens to register any events in their neighborhood like falling trees, holes on the street, and so on.

Technologies: Java EE (EJB, JSF (RichFaces), CDI, Hibernate), JBoss Seam, JUnit, Maven, JBoss AS