Ceron Matthew
Calsena

Archivil-Plan

Built a browser-based construction estimation tool that lets users upload a floor plan image and extract measurable data automatically no manual measuring required. The tool maps the uploaded image into scatter-based coordinate points, uses those coordinates to calculate a Bill of Quantities (BOQ) covering materials such as cement, and compiles everything into a structured Excel report ready for contractor or client handoff. Built entirely with vanilla HTML, CSS, and JavaScript with no backend server all image processing and coordinate mapping runs client-side. Claude AI was used to accelerate development and assist in implementing the coordinate-to-measurement conversion logic.

⚠️ Shared on portfolio with owner's permission.

StreetWise PH

Replaced handwritten stock records with a digital inventory and mini e-commerce system backed by Firebase Realtime Database. Integrated the Firebase REST API for all read and write operations on stock and sales data, enabling live sync across sessions without page refresh. Built sales analysis features that surface performance trends from accumulated transaction data. Developed an automatic invoice generator that produces buyer receipts and downloadable Excel sales reports using SheetJS triggered directly from the sales interface without manual input. Integrated the ImgBB API for product image hosting, connecting image uploads directly to inventory records via REST so product photos are stored and linked automatically. The entire system runs client-side using HTML, CSS, and JavaScript no backend server required. Claude AI was used to accelerate development across the full build.

Booking Record System

Replaced paper booking logs with a calendar-driven reservation system connected to Firebase Realtime Database via REST API. The system handles all reservation CRUD operations through the Firebase REST API, keeping data consistent across multiple devices in real time. Built automated conflict detection that queries Firebase on every booking attempt and blocks overlapping reservations before they are saved. Added an analytics module that reads accumulated booking data from Firebase and surfaces patterns such as peak dates and occupancy trends giving the resort owner data-driven visibility into availability and pricing decisions. The full system runs client-side using HTML, CSS, and JavaScript with no backend server. Claude AI was used throughout development to accelerate delivery.

⚠️ shared on portfolio with owner's permission.

Auto Mailer

Automated repetitive daily email tasks end-to-end by integrating the EmailJS API and Firebase REST API in a single browser-based interface. Firebase supplies dynamic template data recipient details, subject content, and attachment references and passes the payload directly to EmailJS, which handles programmatic mail dispatch. The two APIs work in tandem so the entire send workflow executes from a single button click without opening Gmail or switching browser tabs. Built conditional JavaScript logic for automatic file and image attachment based on user selections, ensuring the correct files are always attached without manual intervention. The system also supports bulk sends with templated content, allowing multi-recipient email campaigns to be triggered in a single action. Built entirely with HTML, CSS, and JavaScript no backend required. Claude AI was used to accelerate development and integration work.

⚠️ shared on portfolio with owner's permission.

PDF Automator

Engineered a browser-based automation app that handles simultaneous import and processing of up to 2,000 PDF files entirely client-side with zero server-side processing. Uses the PDF.js library to parse raw PDF binary streams directly in the browser, extracting text content from each file without uploading anything to a server keeping all data local to the user's machine and ensuring full privacy compliance. Built 3 extraction modes Multiple Keywords, Single Keyword, and Extract All each triggering a distinct automated pipeline tailored to different review workflows. All extracted data is exported directly to Excel via the SheetJS library, producing a structured report without any manual copy-paste. The application runs entirely in HTML, CSS, and JavaScript. Claude AI was used to accelerate development and assist in implementing the PDF parsing and extraction pipeline.

⚠️ shared on portfolio with owner's permission.

Slider Data Mapping

Built a C# WinForms desktop application that reads raw Excel files and flat data exports from manufacturing equipment and converts them into clear, interactive visual charts. The tool parses Excel files directly using file I/O and data mapping logic built in C#, transforms the raw row data into structured chart datasets, and renders them as interactive graphs inside the WinForms interface. Supports export to both Excel and PDF so engineers can share findings without additional tools. Developed as a special internal project at Western Digital to give the engineering team a faster way to interpret machine-generated datasets and identify patterns, anomalies, and root causes that are difficult to spot in raw spreadsheet form.