Hi I'm Tabitha S- I build software that you need.

github

Tabitha Sulaiman  • Peerlist

Built a Python-based import/export pipeline to process Baselight export data and Xytech work orders, transforming local filesystem paths into standardized facility storage locations. Implemented parsing logic to clean and normalize inconsistent path structures, accurately map file locations between systems, and generate a structured CSV output for downstream production use. Designed frame-processing functionality to consolidate consecutive frame numbers into clear ranges while preserving separation by source location. This improved readability and ensured accurate tracking of frames requiring fixes. The solution reduced manual reconciliation effort, streamlined handoffs between color and production teams, and established a repeatable workflow for converting raw text-based data into actionable outputs.

Baselight to Xytech Frame Mapping Script

CSUNNav | Classroom Navigator

Proposed and led the end-to-end development of a centralized website for the CSUN National Society of Black Engineers to improve visibility, member engagement, and access to resources. Defined project scope, created timelines, secured a domain, and coordinated development efforts to ensure structured execution. Designed the system architecture and user flows, translating organizational needs into a scalable multi-page application with consistent navigation, dropdown menus, and integrated donation functionality. Engineered a responsive frontend using React, implementing reusable components, shared layouts, and client-side routing to improve maintainability and user experience. Built interactive features including live newsletter polling and curated internship updates using JavaScript, enabling real-time engagement and reducing reliance on manual communication channels. Leveraged Figma to prototype UI/UX designs and guide development decisions, ensuring a cohesive and intuitive interface across pages. Delivered a modular, production-ready platform that streamlined information distribution and improved accessibility for current and prospective members.

CSUN-NSBE Website

Developed a standalone Python command-line tool to automate image processing workflows using argparse and Pillow, with support for watermarking, thumbnail generation, GIF creation, and metadata export. Built the script to handle both individual files and folder-based batch processing, while enforcing consistent ownership-based naming and automatic versioning for every generated asset. Designed modular processing functions for each operation and implemented validation, output directory management, and reusable file naming logic to keep the workflow reliable and organized. The project reduced repetitive manual editing work by turning multi-step media tasks into simple terminal commands, making it faster to generate branded assets, preview thumbnails, and export documentation for processed files.

FFmpeg Image Processing Utility

During a class project, we were tasked with developing an innovative solution, and after researching the topic and recalling an idea from a final-round interview with a company, I suggested we pursue a VR training model for medical students. The company had highlighted this technology as a promising advancement in healthcare, and I immediately recognized its potential to address a major problem: the high costs associated with training medical students using physical tools. Our VR system allowed students to practice procedures in a realistic virtual environment, eliminating issues like "we’re using this device to train someone, so we can’t treat you right now" or "we ran out of tools to train you." Using affordable VR headsets and remotes, we created a cost-effective and accessible solution compared to traditional medical training tools. I also felt personally connected to the project, as my partner had recently undergone knee removal surgery, so we focused our simulation on the knee. As Project Manager and FullStack Developer at CSUN, I led the development of the VR knee surgery simulation in Unity, incorporating inverse kinematics (IK), haptics, and motion tracking for realistic tool interactions. I designed a grading system to assess incision accuracy, pressure control, and tool stability, using procedural scoring for precise evaluation. Additionally, I optimized 3D assets with LOD, GPU instancing, and physics-driven deformation for smooth real-time performance. This project not only gave me valuable hands-on experience in VR development but also allowed me to contribute to a solution that could make medical training more accessible, affordable, and efficient.

MediTrain | VR Medical Trainer

Developed a full-stack task management application that helps users organize projects and tasks in a secure, streamlined workflow. Engineered the backend with ASP.NET Core 8 Web API, JWT authentication, ASP.NET Identity, and SQLite, while integrating an Angular frontend served directly from the API for a unified deployment. Implemented user registration, login, and protected routes to ensure each user could securely manage only their own data. Designed and built core project and task management features, including project creation, task organization, and user-scoped data access. Structured the application for public cloud deployment with Docker and Render, improving portability and simplifying release management. By replacing manual tracking with a centralized web interface, the project reduced friction in day-to-day task management and created a more efficient, scalable workflow for individual users.

Task Tracker | Angular Project

Developed an end-to-end software pipeline to automate media workflows, processing Baselight and Xytech data into a structured SQLite database and generating user-ready reports. Built backend systems in Python with CLI tools (argparse) to support scalable and flexible execution. Engineered video processing workflows using FFmpeg and FFprobe to convert frame data into timecodes, extract clips, generate thumbnails, and apply watermarking. Designed automated Excel reports using OpenPyXL with embedded visuals to improve usability and speed up asset review. Integrated the Vimeo API to upload media and generate CSV reports with video metadata, enabling tracking and distribution. Managed the project lifecycle from requirements to debugging, focusing on modular design, reliability, and efficient data pipelines.

The Crucible | Automated Video Processing & M

Developed a Python-based data pipeline and reporting tool to parse QA exports and load them into a local MongoDB database, enabling faster access to bug data across multiple test cycles. Built the script with argparse, pandas, and PyMongo to import Spring and Fall dump files into separate collections, normalize inconsistent fields such as build dates and boolean values, and generate deduplicated reports for key workflows including test-owner activity, repeatable bugs, blocker bugs, and build-specific issues. From a software engineering and product execution standpoint, I designed the reporting flow around real QA use cases so the tool could turn raw exports into actionable outputs with a single command. I implemented duplicate handling, structured query logic, and CSV export for stakeholder-ready reporting, which reduced manual spreadsheet review and made bug trends easier to track across collections. The command-line interface also reflected strong frontend awareness by prioritizing clear input options, predictable report flags, and readable console output as part of the overall user experience for technical users. This project also strengthened my project management skills by translating assignment requirements into a clean delivery plan: database setup, import workflow, deduplication rules, report generation, and CSV export. I scoped the work around concrete deliverables, validated outputs across multiple report scenarios, and organized the implementation so it could be extended for additional filters, dashboards, or internal QA tooling.

The Reckoning | QA Bug Reporting Database Imp

Proposed the concept for a real-time pose detection fitness app during a Shark Tank-style initiative at my internship, then continued development at CSUN by supporting both product direction and technical implementation. Contributed to building an on-device pipeline using Kotlin, CameraX, and TensorFlow Lite MoveNet, incorporating One-Euro smoothing and a finite-state rep detection system to deliver accurate, low-latency feedback. Achieved over 30 FPS and sub-80 ms performance in live sessions without relying on cloud processing. Collaborated on UI/UX design and implementation using Figma and Jetpack Compose, focusing on clear, real-time guidance through color-coded limb tracking, two-cue overlays, and concise microcopy. Helped shape an intuitive user experience that supported smooth onboarding and effective in-session feedback during testing. Supported the development of an offline-first data layer using Room and WorkManager with Supabase, enabling reliable data persistence and sync. Assisted with integrating Google Sign-In and Firebase Crashlytics to ensure secure authentication and stable performance, resulting in consistent session tracking and zero crashes during extended testing.

trAIn | Mobile AI Workout Trainer

Crafted a cross-platform desktop IRC client using Rust, combining Tauri for the native application layer and Leptos for a reactive WebAssembly-based UI. Led the UI design and implementation, shaping an intuitive and responsive interface that integrates seamlessly with a Rust backend. Contributed to both a Tauri-based desktop version and a Godot-powered variant, maintaining a consistent user experience across different rendering environments. Designed interactive chat components including channel navigation, message streams, and real-time updates, focusing on clarity and usability. Bridged frontend systems with asynchronous Rust services through an event-driven architecture, enabling smooth, low-latency communication between the interface and backend. Helped structure a maintainable UI layer that simplifies state handling and improves responsiveness. Delivered a cohesive and performant chat experience that enhances real-time communication workflows. The dual implementation across Tauri and Godot highlights versatility in frontend engineering and the ability to connect modern UI frameworks with systems-level Rust applications.

VeryChat | Rust-Powered Desktop IRC Client (T

Originally called "Procrastination Killah," this hackathon project was inspired by a concept I encountered in class, where we studied the interval scheduling problem. In interval scheduling, we often work with tasks that have start and finish times, and the goal is to efficiently schedule them based on those two time values. This concept inspired our project, though we applied it differently. Instead of just sorting based on start and finish times, we had to sort a list of tasks with three separate values: urgency, estimated study time, and importance. To prioritize tasks, we created an equation that combined these three values into one sortable metric, ensuring that students could focus on the most important and time-sensitive work first. The platform was designed to help students who often feel overwhelmed by their procrastination and don’t know where to start. As Project Manager and FullStack Developer, I optimized academic planning by integrating OpenAI and Firebase to autofill coursework, estimate study time, and prioritize tasks using Timsort. I coordinated the team through Jira, ensuring tasks were completed on time and the platform was delivered as planned. I also designed the UI with React.js and Node.js, incorporating dynamic components, state management, and a drag-and-drop calendar for efficient study planning. This project allowed me to blend problem-solving with technology to create a tool that helps students stay organized, manage their time effectively, and reduce stress.