Hey, I'm Roy.

I'm a |

At BNY, I design and deploy enterprise ML/AI systems that help 1,400+ people work smarter, from agents automate financial regulatory reporting to assistants that help with faster data discovery. I'm passionate about applying data science to solve complex business challenges. Feel free to explore my work and reach out!

BNY NYC Open to collaboration
RZ

Experience

Where I've Worked
Data Scientist @ BNY
August 2023 - Present
  • Designed and deployed enterprise AI/ML solutions for the bank's data platform.
  • Built an agent integrated with a larger digital employee to extract and standardize data from unstructured financial documents, achieving 96% extraction accuracy and reducing manual processing to allow teams to focus on analytics and strategic work.
  • Led development of an AI-assistant trained on technical documentation and enterprise knowledge sources to deliver role-aware, context-sensitive support across products and features for 1,400+ employees, improving self-service, reducing support load, and accelerating data onboarding.
  • Engineered the company's first regulatory agent for reviewing supervisory findings, auditing remediation progress, handling CSI securely, and drafting responses, successfully addressing 15+ findings from regulators.
  • Developed a data governance tool that orchestrated multiple agents to automate generation of data dictionaries, standardized business terms, and data quality rules to enhance metadata accuracy, enforce a single source of truth, and improve enterprise data quality.
  • Developed a RoBERTa-based sequence classification model to assign sensitivity levels to data lake elements, leveraging NLP to improve compliance for over 2000 datasets.
  • Implemented an ARIMA-based time series anomaly detection model to identify abnormal data ingestion patterns in the data platform, proactively detecting pipeline issues and supporting operational monitoring.
  • Built statistical and machine learning models to automate SLA assignment for data providers, improving operational efficiency and resulting in ~25% fewer late deliverables.
Python SQL Snowflake pandas scikit-learn XGBoost RAG AI Agents LLMs
Web Scraping Specialist @ Octavate
March 2023 - April 2023
  • Built and maintained automated web scraping pipelines, storing datasets in AWS, and processing them to extract insights to support decision making.
  • Partnered with analytics team to drive discovery and integration of new data sources, expanding the company's data capabilities.
Python SQL BeautifulSoup Selenium AWS
Software Developer Intern @ Instinet
June 2022 - August 2022
  • Developed a client application interfacing with the company's trading platforms to submit orders, execute trades, and consume real-time market data.
  • Consolidated end-to-end testing and validation of legacy and modern FIX-based trading platforms.
  • Collaborated with senior engineers in an Agile environment to deliver features in two-week sprints while gaining experience with financial systems, communication protocols, and enterprise codebases.
Java Jakarta Servlet JUnit Gradle AWS FIX Protocol
Research Intern @ Sagerize
  • Analyzed the Common Data Set of over 100+ colleges and universities around the country and evaluated factors important for admissions.
  • Compiled data for college essays and resumes written by users of the platform.

Projects

MarketLens

MarketLens

Stock research assistant integrating market data and news with AI/ML to provide data-driven investment insights.

Python JavaScript Gemini API PyTorch Transformers (Hugging Face)
Syllabal

Syllabal

Conversation-driven language learning platform using AI companions to simulate real-world scenarios and provide immediate feedback.

Python Next.js LiveKit TypeScript Digital Ocean
Options Price Forecasting

Options Price Forecasting

Deep learning models for predicting option prices.

Python TensorFlow Pandas
Poverty Prediction ADS Audit

Poverty Prediction ADS Audit

Accuracy and fairness analysis of an automated decision system for predicting poverty level.

Python Scikit-learn Pandas
Competitiveness of Local Government Pay

Competitiveness of Local Government Pay

Analysis of local government pay using big data tools and technologies.

Python PySpark Hadoop
Fingertip Position Estimations

Fingertip Position Estimations

Convolutional neural network to estimate the positions of each fingertip on a robotic hand.

Python PyTorch CNN

Let's Connect

I'd love to hear from you.