Construex
Remote | Oct. 2025 - Current
Software Developer
At Construex, I work on AI-powered tooling for the construction materials industry, with a focus on making product discovery and seller workflows smarter and more efficient. My core project has been re-architecting the company’s chatbot from a monolithic LLM approach to a modular, intent-routing pipeline using small language models (SLMs) and vector embeddings — breaking broad prompts into intent-specific chains that reduced AI inference cost by 25% per query while improving response latency on product searches.
Beyond the chatbot, I built a multimodal product categorization system that classifies construction materials from images, designing the evaluation pipelines and iterating on prompts to reach consistent, accurate results. I’m also actively building an agentic pipeline that ingests unstructured spec sheets, extracts and normalizes SKU attributes, and persists searchable records to a database — directly improving both seller onboarding speed and customer search relevance.
Abstrac
Remote | Sept. 2025 - Current
Full Stack Developer
Abstrac is a cross-platform social media app, and I lead its full-stack development end-to-end. The tech stack is built on Expo (React Native) for the client and Firebase on the backend — spanning Authentication, Firestore, Cloud Functions, and Cloud Storage to support real-time feeds and media-heavy workflows.
On the backend, I designed the Cloud Functions API layer, normalized Firestore schemas with composite indexes and security rules, and implemented CI/CD alongside batched reads and query optimizations to keep latency low and cloud costs predictable. On the client side, I built a caching layer in Expo that balances Firebase read costs against perceived performance, significantly cutting redundant fetches on media-heavy screens.
BrickRed Systems
Seattle, WA | Apr. 2025 - Jul. 2025
Software Developer Contractor
At BrickRed Systems I contributed to a multi-service AI document processing platform focused on real estate. My primary work was optimizing the document ingestion pipeline — specifically building an XGBoost + Optuna training workflow to classify and filter non-informative pages before they reached the LLM, cutting Azure OpenAI token costs while improving the reliability of downstream predictions.
I also designed and implemented a CI/CD pipeline in Azure DevOps to automate the build, test, and deployment cycle across the platform’s services, replacing a manual process and giving the team a more consistent and auditable release workflow.
Intel Nonvolatile Memory Storage Group
Folsom, CA | June 2019 - August 2019
Software Developer Intern
I was part of Intel’s NSG Tool Development Group, where my main project was building a Python script to parse a wide variety of XML formats across different NSG products — automating the extraction and analysis of critical information that had previously required manual review. Alongside that, I created and maintained database tables in SQL to support efficient data organization and retrieval for the team’s tooling needs.
SFApex
Houston, TX | June 2018 - August 2018
Software Developer Intern
I joined a development team building a Salesforce app specialized in data migration between sandboxes. Working within MVC design practices, I implemented a Data Compare feature in Node.js that pulled data from DBeaver SQL servers and ran it through a JavaScript comparison algorithm. I also built out data validation for incoming database records using JavaScript’s Joi library to ensure integrity throughout the migration pipeline.