York University, ON

BEng, Software Engineering (Big Data Stream)

GPA: 7.82 GPA (3.75/4.0)

Relevant Courses: Data Structures and Algorithms, Operating System Fundamentals, Software Design, Object Oriented Programming, Database Management Systems, Big Data Systems, Software Engineering Testing

Software Engineer, Bloomberg L.P.

• Built and maintained a petabyte-scale data lake (~840 production tables) built on the Apache Iceberg table format, serving as the storage and query backbone for a financial pricing product
• Developed and operated decoupled maintenance services for the Iceberg tables (compaction, snapshot expiration, orphan-file cleanup, and TTL-based retention) to control storage growth and sustain query performance at scale
• Built and operated a Trino cluster on container-native infrastructure serving tens of thousands of queries/day for 200+ users, delivering interactive, ad-hoc SQL analytics over petabytes of data
• Built a self-service streaming platform on Apache Flink that lets teams onboard Kafka sources into Iceberg, delivering real time data at hundreds of billions of records per day
• Extended the platform with a Spark-based batch ingestion framework that handles high-volume backfills and on-demand loads into Iceberg, abstracting underlying complexity behind declarative config
• Orchestrated data pipelines with Argo Workflows, leveraging it for both scheduling and end-to-end observability into pipeline execution and health

Software Engineering Intern, Meta Platforms, Inc

• Worked on Presto, an open source distributed query engine for fast and easy SQL analytics over big data
• Implemented the catalog server which serves as a distributed cache for metadata to reduce cpu/memory pressure and calls for external dependencies
• Defined and implemented an Apache Thrift service to transport metadata results between servers through serialization/deserialization of the objects. Made use of Jackson Modules to resolve interface implementation types
• Added the catalog server as separate node by injecting dependencies with Google Guice and installing them upon startup. Allowed it to communicate with the presto server and added end to end tests to confirm functionality
• Enabled multi level caching between the presto server and catalog server for metadata results. Added run time stats to monitor performance metrics like cache hit rate
• Did performance testing on a cluster and resulted in a cache hit rate of 65% on metadata calls
• Saw about a 30% improvement on CPU and query execution time on 95th percentile latencies

Software Engineering Intern, Squarepoint Capital LLP

• Deployed captures for feed handlers used to process, aggregate and store market data onto a company server for clients and quants to use
• Modernized and improved two of the companies critical feeds in C++
• Migrated the two feeds publishing protocol to its modern version. This migration increased trades published per symbol by 62.6% in the feeds
• Applied modern sequencing logic in the two feed decoders extending packet gap reports being logged as well as improving message ordering reducing packet gaps by about 10%
• Split both feed decoders into decoder/parser and improved message processing logic increasing conflation on trade messages being generated by 10%
• Used the factory design pattern for decoder/parser creation on the two feeds allowing clients to create feed objects and be introduced to new versions of the feeds without exposing their creation logic

Back End Developer Intern, IBM Canada Limited

• Worked on the Main Data Structures used to Optimize Multi-dimensional Queries for the Calculation Engine in the IBM TM1 Server
• Created Functional and BDD Style Unit Tests for Classes over several Data Types using the Catch2 Framework, Increasing Testing Coverage in Several Classes to Over 90%
• Created Scalable Benchmarks to Characterize Classes Performance over several Data Types using Google Benchmark, Allowing for Baseline Comparisons of Code Changes
• Optimized and Resolved Errors in the Data Structure Used to Insert and Delete Ranges Using Memory Management, Increasing Certain Executions By About 10%

Sales Associate, Canada Computers

• Communicated with customers in person to find their specific needs and answer technical questions
• Ability to multitask and pick up skills (was able to work as the cashier and merchandiser)
• Knowledge on computers, laptops, networking, and accessories within the store

SpaceDefender (Eiffel)

• Implemented the business model from the specification of an abstract UI for an RPG game
• Created acceptance tests for my business model while performing regression testing while developing it
• Used the state design pattern to implement the moving and specific details of different states in the game
• Made a design diagram specifying the client/supplier and parent/child relationships within the project

Project Repository: github.com/erick576/SpaceDefender

VennFX (Java, JavaFX, MySQL)

• An Interactive Venn Diagram GUI Application that can Organize and Import/Export Information
• Used Gradle to manage Dependencies as well as Automate the Building and Testing of the Application
• Implemented Undo/Redo Functionality from scratch by creating an Interface that Abstracted different types of Undo/Redo Commands

Project Repository: github.com/erick576/VennFX

StudentRecordSystem (Java, JavaFX)

• A GUI Application that is capable of Effectively Storing many Students with their Unique Attributes
• Used Java's JavaFX Framework to implement the Model-View-Controller Design Pattern
• Users are able to Add, View and Edit Student information as well as Export the Data to a CSV File
• Utilized core concepts in Object Oriented Programming (Inheritance, Aggregation, Exception Handling)

Project Repository: github.com/erick576/StudentRecordSystem

MyItemLog (Python, Django, PostgreSQL, HTML, CSS)

• A Django Web Application where Users can keep a Log of all their Possessions (Add, Edit, Delete, Export)
• Utilized a RESTFUL API to create Communication between the Front and Back End of the Application
• Deployed the Web Application on Heroku and hosted the Database on PostgreSQL

Project Repository: github.com/erick576/MyItemLog

WorldPopulationVisualization (Python, Jupyter Notebook)

• Used Python's Framework Jupyter Notebook to Visualize Population Data for Countries around the World
• Scraped Data from an Online Database into a DataFrame using BeautifulSoup and Pandas Libraries
• Used Matplotlib, Seaborn and Plotly Libraries in order to Represent and Manipulate the Information in the DataFrame

Project Repository: github.com/erick576/WorldPopulationVisualization

ConsoleTracker (Python, Django)

• A software tool used to aid existing software in gaming cafes that cant exercise the same level of control over consoles such as Xbox, PlayStation and other gaming devices
• Implemented a socket listener running as a background process to listen to requests from an external client
• Handled requests to either update a users time balance or query all of the machine statuses
• Extensively tested entities in the system using equivalence and boundary value testing on each of their values

Project Repository: github.com/erick576/ConsoleTracker