Marco Scarlata



Skills

  • Languages: C++, Python, Java, JavaScript, HTML/CSS, SQL

  • Frameworks: gRPC, Protocol Buffers, Flask, Django, React, Svelte, Bulma

  • Libraries: TFLite, Node.js, Selenium, SpaCy, Keras, Pandas, NumPy, NLTK

  • Tools: Build, Bazel/Blaze, BigQuery, Bash, Git, Kafka, DynamoDB

Current Learning Overview

Education

University of Rochester

BA in Computer Science Aug 2018 – May 2022


  • Major GPA: 3.5/4.0 – Consecutive Dean’s List Recipient
  • Relevant Courses: Data Structures & Algorithms | Formal Systems & Computation | Web Programming | Human Computer Interaction | Intro to Artificial Intelligence | Database Systems | Natural Language Processing
  • Dean's Scholarship & Rochester National Grant Awardee

Experience

Google Software Engineer

San Francisco, CA Sep 2022 - Present

Pixel Watch Device Algorithms


  • Partnered with a Research Scientist to develop end-to-end, a real-time gesture recognition algorithm using C++, TFLite, and Python, integrated with various IMU sensors and optimized for low latency and power consumption.
  • Augmented and took code ownership of the Low Latency Off-body Detection Sensor algorithm using C++.
  • Refined core algorithm components (e.g., data buffering, sensor fusion), and designed telemetry systems with metrics dashboards and analysis pipelines, resolving 50+ bugs to improve accuracy and latency.
  • Collaborated with cross-functional teams to integrate the gesture-based algorithm into 1P apps, resolving system-level conflicts and enabling new user interaction methods, culminating in recognition through high-visibility demos.

Cloud Asset Inventory & Search


  • Productionized a reliable, scalable solution in C++ to enrich Cloud Asset Inventory with structured, enhanced metadata into the centralized data warehouse for cloud-managed assets, impacting over 90% of GCP Resources and increasing user adoption YoY with a current user base of over 500,000 active users.
  • Assisted in launching Asset Enrichment end-to-end into the Asset Query System using C++, SQL, and Spanner, leveraging enriched metadata to deliver efficient query results of more than 275 GCP asset types for SCC customers.
  • Drove a cross-team engagement to optimize a workflow runner using gRPC, Borg, and Python, reducing development time by 25% and saving the equivalent of ~10 SWE weeks.

OpenSesame Software Engineering Intern

Portland, OR Jun 2020 – Aug 2020

Front-End Division


  • Extended language drop-down feature to include search selection for courses through Angular and TypeScript
  • Extrapolated multiple burn-down processes with Behat, API and Unit tests resulting in locating over 6 different bugs spawned from the most recent sprints, resolving the bug which prevented the landing page from loading for IE users
  • Built out E2E tests through replicating over 10 deprecated legacy tests from Drupal to Angular & Selenium

University of Rochester Teaching Assistant & CETL Tutor

Rochester, NY Sep 2019 - Dec 2021


  • Analyzed the purchase history of 100,000+ users buying coupons by regression analysis using Python, Pandas, NumPy and PyTorch, thereby locating an under-marketed sector in sales that led to a 10% increase in revenue
  • Launched an MVP module-based recommender system with caching for an ecommerce platform called BananaMall with 10,000+ downloads on Playstore and 100,000+ users using Python, SKLearn and DynamoDB

ANDSystems Machine Learning Intern

Ulaanbaatar Mongolia Jun 2019 – Aug 2019

Machine Learning Team


  • Analyzed the purchase history of 100,000+ users buying coupons by regression analysis using Python, Pandas, NumPy and PyTorch, thereby locating an under-marketed sector in sales that led to a 10% increase in revenue
  • Launched an MVP module-based recommender system with caching for an ecommerce platform called BananaMall with 10,000+ downloads on Playstore and 100,000+ users using Python, SKLearn and DynamoDB

Projects

Trading Algorithm Tensorflow, Python, MySQL


  • Developed a mock quantitative investment platform that utilizes Tensorflow to deliver accurate trading predictions
  • Includes Python quantitative trading strategies including Pattern Recognition, Commodity Trading Advisor, Monte Carlo, Options Straddle, Heikin-Ashi, Pair Trading and VIX Calculator