Intro
I am a systems software engineer focused on multimedia machine learning systems, and efficient storage/retrieval. I aspire to spend the rest of my career working with low-level C++ and ML systems. I am looking to further develop these skills through full-time employment.
Prior to my journey into software engineering through Arduino projects, I worked as a Medical Assistant in healthcare. This experience provided me with a strong foundation in clear communication and collaborative teamwork, skills that I have found translate seamlessly into effective software development.
At the present moment, this page is under construction. Soon, it will track my technical tutorials, art and music! Stay tuned for updates!
Education
Georgia Institute of Technology
May 2027M.S, Computer Science
May 2027
Coursework:
- Advanced Operating Systems, Foundations of Computer Graphics
New York University
Dec 2024Master of Science, Computer Engineering
Dec 2024
Coursework:
- Discrete Mathematics, Operating Systems, Computer Networking, Computer Architecture, Internet Architecture and Protocols, Data Structures & Algorithms
- Machine Learning, Deep Learning, High Performance Machine Learning, Signal Processing
Certifications/Awards:
- Bridge to Tandon, Merit Scholarship
- Amazon Web Services: Solutions Architect
Activities:
- Society for Hispanic Professional Engineers - NYU
- ColorStack NYC
Bachelor of Science in Biochemistry
May 2021Binghamton University
May 2021
Certifications/Awards:
- Dean's List
Activities:
- Piano Society of BU: Event Coordinator and Piano Instructor (2018-2020)
- Volunteer at Lourdes Hospital - Emergency Room (2019-2020)
- Big Brother Big Sister - Twin Cities (2018-2020)
Experience
Machine Learning Engineer
May 2025 - PresentVisentra
New York, NY
May 2025 - Present
Responsibilities:
- Architected an image & video processing pipeline using AWS Lambda, Step Functions, and EventBridge with Golang.
- Implemented a dual-mode video super-resolution system using AWS: (1) AWS Batch with GPU instances for immediate processing, and (2) Amazon SQS with Spot Instances and Kubernetes for cost-effective queued processing.
- Deployed super-resolution and frame-interpolation models on AWS SageMaker under an EKS-managed inference service with ONNX Runtime, scaling requests dynamically via an API-driven workflow that improved throughput by 2.2x.
Technologies:
Python
Go
Pytorch
ONNX
TensorRT
AWS
KubeFlow
WandB
Kubernetes
Docker
Community Data Science Intern
Jan 2025 – April 2025BroadStreet
Remote
Jan 2025 – April 2025
Responsibilities:
- Built SQL pipelines ingesting and joining NYC DOHMH Restaurant Inspection Results with Restaurant Grades, normalizing millions of records to track violations and food safety outcomes across 25,000+ establishments
- Developed geospatial heatmaps and interactive dashboards (PostGIS, Python, Plotly) to visualize borough-level hotspots of critical violations and grade distributions over time
Technologies:
Python
Go
Pytorch
ONNX
TensorRT
AWS
Kubernetes
Software Engineering Intern
Sep 2024 – Dec 2024Sonara
New York, NY
Sep 2024 – Dec 2024
Responsibilities:
- Architected several scalable AWS-based audio-processing APIs using ECS Fargate, SQS, and Step Functions with Golang.
- Optimized DSP algorithms, including filters, EQ, reverb, and compression, using C++/MATLAB and ffmpeg for real-time audio processing.
- Enhanced computational efficiency and performance to achieve high-quality sound.
Technologies:
Python
Go
Pytorch
ONNX
TensorRT
AWS
Kubernetes
Research
Machine Learning Research Intern
May 2024 – Aug 2024Nagoya University, Signal Processing Lab
Nagoya, Japan
May 2024 – Aug 2024
Responsibilities:
- Investigated super resolution model DRCT’s limitations and identified potential bottlenecks for deeper architectures; proposed modifications inspired by SwinV2, including cosine attention, post-normalization, and continuous log-space relative bias.
- Implemented and trained modified architectures on the DIV2K dataset using PyTorch with mixed-precision training, cosine learning rate scheduling, and distributed data parallelism (A100); validated across 2×/4× upscaling benchmarks with PSNR.
- Built comprehensive ablation framework with automated benchmarking across 5 model variants to evaluate the trade-offs between architectural complexity and output fidelity, revealing 28.2 - 28.6 PSNR variance with 25% less memory overhead
Biochemistry Computational Research
Feb 2019 – Apr 2020Innovation Technologies Complex, Catalysis Lab
New York, NY
Feb 2019 – Apr 2020
Responsibilities:
- Developed molecular dynamics simulations in Python/C++ to model Sonic Hedgehog Protein (SHH) self-catalyzing cholesterolysis, analyzing free energy landscapes and reaction pathways to understand catalytic efficiency
- Implemented reproducible computational workflows integrating docking, QM/MM calculations, and custom visualization scripts to quantify cholesterol–protein interactions and validate hypotheses against experimental kinetics data
C++
Python
CUDA
SQL
PyTorch
AWS
gRPC
Docker
Kubernetes
DynamoDB
Redis
SDL2
ffpmeg
ONNX
TensorRT
Databases
Redis - K/V Database- C++
A simplified Redis-like in-memory data structure store, capable of supporting basic data structures such as strings, hashes, lists, sets, and more, with a command parser to interpret and execute basic Redis commands.
Recent Blogs
Reading List
Reading List
Bjarne Stroustrup
A crash-course on the foundations of C++ written by the language author Bjarne Stroustrup.
Notes Here
