Problem
Personal project
AFA Text-to-Speech
A cross-platform floating desktop app that turns typed text into speech and routes the audio into meeting applications as virtual microphone input.
ElectronJavaScriptSwiftC#Linux Audio
What it does
- Always-on-top desktop speaking interface
- Virtual audio routing through BlackHole, VB-CABLE, PipeWire, and PulseAudio
- Packaged Windows and Linux builds with diagnostics and setup flows
AF
Inside the build
Problem, constraints, decisions, and outcome.
What changed and why
Constraint
Virtual-audio installation and routing differ substantially across macOS, Windows, and Linux, and packaging must not promise drivers that are absent.
My contribution
Built the floating desktop interface, shared speech controller, platform-specific routing and driver checks, diagnostics, onboarding, and distributable builds.
Technical decision
Separate speech generation from audio routing and use each platform's established virtual device layer: BlackHole, VB-CABLE, or PipeWire/PulseAudio.
Outcome
Produced shareable Windows and Linux builds with setup diagnostics, while keeping unfinished platform capabilities explicit rather than hiding them.
More Work
Other projects from the same portfolio.
Geez Document Retrieval System
Dedicated Information Retrieval system for Ge'ez texts, featuring a custom stemmer and evaluated using precision, recall, and F1-score.
Civic Reporting System
Location-based civic reporting platform that clusters community issues and routes them to relevant authorities.
Raster Graphics Simulator
Sandbox environment for experimenting with raster graphics algorithms and rendering trade-offs.
HaFCoM Corporate Website
Official website for an Ethiopian construction and architectural firm.
Fast Food Restaurant Management System
Queue-based system for coordinating kitchen workflows in fast-paced restaurants.
CPU Scheduling Algorithm Visualizer
Interactive simulator for FCFS, SJF, and Round Robin scheduling algorithms.