Gabriel Nieves

August 2017 – May 2020

Machine Learning & AI

• Information Retrieval
• Deep Learning
• Natural Language Processing
• Feature Engineering

Distributed Computing

• RESTful Web Services
• Containerization & Deployment
• MapReduce

July 2009 – July 2017

Computer Science

• Algorithms
• Data Structures
• Operating Systems
• Databases

Computer Engineering

• Computer Architecture
• Logic Circuits

Electrical Engineering

• Digital Signal Processing (DSP)

Mathematics

• Linear Algebra
• Calculus
• Differential Equations
• Probability and Statistics

Business & Economics

• Entrepreneurship
• Engineering Economics

November 2021 – Present

As a Senior Data Scientist, I lead customer-facing initiatives, encompassing the ownership, execution, and design of tailored solutions for complex, fast-moving challenges.

Key Accomplishments

• Co-developer of LazyGraphRAG, a lightweight, cost-effective GraphRAG solution: Microsoft Research Blog

• Collaborated in the development and open-source release of the Microsoft GraphRAG library: github.com/microsoft/graphrag

• Co-led the design and development of the GraphRAG Accelerator v2.0.0 API, culminating in its open-source release: github.com/azure-samples/graphrag-accelerator

• Developed a stochastic gradient descent–inspired data triage algorithm that automates dataset attribute parameterization, embedding generation, clustering, and user-defined configuration pipelines. Delivered as a Python package with a Bash CLI.

• Developed a scalable multi-threaded Python interface for SQL databases utilizing linear algebra algorithms, image masking techniques, and object-oriented design. The solution achieved an order-of-magnitude throughput improvement and contributed to the extension of a strategic customer engagement.

August 2020 – October 2021

• Responsible for data engineering (ETL), health monitoring, and cloud analytics using Java MapReduce.
• Conducted feature analysis for optimized data ingestion.
• Developed an ingest metrics service for health monitoring and predictive trend analysis.

January 2019 – July 2020

• Developed a scalable semantic image retrieval system for petabyte-scale datasets.
• Researched image feature extraction and indexing methodologies.
• Built a testing platform for evaluating storage architectures and retrieval algorithms.
• Helped characterize speed/accuracy tradeoffs for production research systems.
• Created a Python package for serialization/deserialization of quantized image feature vectors.
• Retrained and fine-tuned ResNet-50, DenseNet-169, and Inception-v3 using multiple open-source datasets.
• Conducted performance research on distributed filesystem I/O, uncovered utilization at just 20% of theoretical capacity, and developed caching and data-distribution improvements that increased read performance by 5×.

• Python, OpenCV, C++, Java, LaTeX
• Hadoop, MapReduce, Docker
• MySQL, NoSQL
• Microsoft Azure, Amazon Web Services (AWS)
• Jupyter Notebooks
• Computer Vision, Machine Learning, Deep Learning
• Fine-Tuning, Large Language Models (LLMs)

• Introduction to Statistical Learning with PyTorch — Udacity Nanodegree (Nov 2020 – Jun 2021)
• Deep Learning Nanodegree — Udacity (Jan 2019 – Jun 2019)
• Introduction to Machine Learning — Stanford Online, Coursera (Jan 2018 – Apr 2018)