Workshops, student research projects, and open-source systems through IEEE SPS at UF, alongside the apps and tools I build independently: Ora, Gradus, and Media Cull Suite.
The teaching library behind the IEEE Signal Processing Society chapter at UF: 87 Jupyter workshops in 12 topics, cut into 30-40 minute sessions, with key numerical claims in the 70 verified workshops checked against an independent oracle.
The chapter's public website: a React/TypeScript/Vite single-page app with typed data modules for events, projects, and the board, a research status page, event RSVPs, and a GitHub Pages deploy pipeline.
Aude develops an end-to-end audio research workflow: collect synchronized microphone-array data, benchmark state-of-the-art baselines, and train models for robust source separation and localization.
System that extracts structured nodes (claims, evidence, limitations, etc.) from papers and scores them to support literature review and comparison across a collection.
Ergo integrates biosignal hardware, feature extraction, and simulation to test how control systems move between stable and unstable regimes under cooperating vs. competing subjects and fatigue.
Robotics project focused on building a reliable hardware-to-software interface layer (instrumentation, control, and real-time data paths), with planned integration points for Ergo.
A co-founded, local-first training and progress app for lifters: on-device logging, optional cloud or on-device language models under strict contracts, and a biomechanics research spine that says what it will not claim.
Gradus turns rigorous STEM courses into concept-graph practice: deterministic grading, a diagnosis of why an answer missed, a repair target chosen by walking the prerequisite graph, and a transfer check before moving on.
A local-first culling suite that ranks a shoot, learns a photographer's taste, and closes a capture loop back to the camera so the next shoot comes out better, without uploading a single frame.
Vie investigates real-time machine perception systems that jointly model moving and static objects, with reproducible data engineering and demo-first iteration.