Open to new opportunities

Roberto Lozano Jr.

Full Stack Developer

I build React-based production systems, and the agentic workflows that ship them faster.

3+ years of professional experience building full-stack web applications with React, Node.js, and PostgreSQL. MS in Computer Science with graduate research in NLP and machine learning.

View Resume Get in Touch

3+: Years professional
MS: Computer Science
10: Shipped projects

Experience

Where I've worked

Production systems and internal tooling for construction and manufacturing teams.

Software Developer

IntraSync Industrial · Contract

Dec 2025 - Jul 2026Remote

Built and maintained a full-stack web portal managing the end-to-end construction panel pipeline spanning estimating, production, inventory, shipping, and accounting.
Implemented WebSocket-based real-time features to keep data in sync across concurrent user sessions.
Managed deployment infrastructure with Docker and Portainer, maintaining containerized API and web service builds on Azure DevOps.
Built custom Claude Code subagents with defined roles, tool scopes, and isolated context windows for code review, test authoring, and codebase exploration.
Ran parallel agents across independent workstreams on isolated git worktrees and verified UI changes through Playwright browser automation, reviewing all generated code before merge.
Collaborated directly with clients to gather requirements, surface edge cases, and iterate on features from real-world feedback.

React
FeathersJS
Node.js
PostgreSQL
WebSockets
Docker
Portainer
Azure DevOps
Claude Code
Playwright

Full Stack Software Developer

Walters & Wolf

Aug 2021 - Feb 2024Fremont, CA

Led development of React-based internal web tools, independently managing front-end systems while contributing to PostgreSQL schema changes and API routing.
Built and maintained shared component libraries, improving reusability across teams and reducing developer onboarding time.
Introduced Storybook documentation for React component libraries and implemented Jest unit tests, improving code quality and team efficiency.
Led stakeholder communications, gathered feedback, and delivered prioritized updates to keep tools aligned with business needs.
Collaborated on Azure DevOps CI/CD workflows, ensuring reliable deployment pipelines and version control integrity.

React
JavaScript
MobX
Node.js
PostgreSQL
Storybook
Jest
Azure DevOps

Education

Where I studied

An MS focused on machine learning and NLP, on top of a CS foundation.

Graduate coursework concentrated in machine learning, NLP, and information retrieval, alongside systems and cloud architecture. Research projects covered word embeddings, recommendation systems, and sentiment analysis.

Selected Coursework

Natural Language Processing
Deep Learning
Machine Learning
Artificial Intelligence
Pattern Recognition and Data Mining
+7 more

Broad CS foundation across algorithms, systems, databases, and networks, with upper-division electives in machine learning, computational linguistics, and web programming. Interactive media design coursework produced the first projects I shipped publicly.

Selected Coursework

Web Programming
Machine Learning
Database Systems
Software Engineering
Computational Linguistics
+8 more

See full coursework

Projects

What I've been up to

Selected personal and graduate research work across web, ML, and mobile.

Web App

Hungr

Hungr is a web app that solves the problem of not being able to decide what to eat. Similar to the tinder style of swiping, users will swipe on which restaurants they like to decide on a winner.

HTML
CSS
Javascript
Node
Yelp API
+3

Code Live Site

Machine Learning

Improving Medical Terminology With Word Embeddings

Improved English-to-Spanish medical term translations by aligning FastText embeddings using Orthogonal Procrustes Analysis. Built a bilingual dictionary and evaluated with BLEU scores, outperforming Google Translate on semantic accuracy, then validated the alignment visually with t-SNE.

Python
FastText
NumPy
t-SNE
BLEU

Code

Machine Learning

Sentiment Analysis with Neural Networks and TF-IDF

Trained a PyTorch-based sentiment classifier using TF-IDF vectorization and k-fold cross-validation. Preprocessing included bi-grams, stop-word removal, and dropout regularization.

Python
PyTorch
TF-IDF
scikit-learn

Code

Machine Learning

Collaborative Filtering Recommendation System

Built a hybrid recommendation engine using user-based, item-based, and variance-based collaborative filtering. Applied Inverse User Frequency and variance adjustments to improve accuracy on sparse datasets, with Cosine, Pearson, and Adjusted Cosine similarity metrics combined into an ensemble.

Python
NumPy

Code

Machine Learning

Probabilistic Latent Semantic Analysis Implementation

Implemented the EM algorithm to extract topics from a term-document matrix using PLSA. Built custom preprocessing pipelines and validated insights through log-likelihood convergence.