Isabella Wu (wuisabel-gif)
Isabella Wu is an individual student developer and open-source creator, not an incorporated company. She builds practical, privacy-conscious software and hardware tools spanning embedded systems, robotics, localization, AI-agent tooling, and creative AI, with an emphasis on making low-level or difficult workflows more accessible.
At a Glance
- Developers and coding-agent users
- Robotics and autonomous-vehicle teams
- Embedded-systems and avionics developers
- AI-video creators, fashion filmmakers and visual storytellers
- +1 more
AI Tools by Isabella Wu (wuisabel-gif)
(1)
MemoryWhale
Local Debug Memory for Agents
Discussions
No discussions yet
Be the first to start a discussion about Isabella Wu (wuisabel-gif)
Latest News
WireDAQ continued active development with new repository commits to its DAQ simulation and test infrastructure.
Cadence repository recorded its latest development milestone for the writing de-slop and voice tool.
MemoryWhale 0.10.0 released as Agent-Native Memory, adding Claude Code/Rho adapters, explicit agent metadata, repository identity and local HTTP interfaces.
Delphin repository recorded a latest commit for the Rust duplex AI-agent CLI companion.
Products & Services
Rust/Tauri local-first terminal memory system for developers and coding agents. It records commands, output, errors and fixes into local SQLite and exposes retrieval through CLI, TUI, web/desktop interfaces and MCP; it supports Claude Code, Rho, Codex, Cursor and other compatible clients.
Open-source writing-quality and voice tool that detects prose that sounds robotic, explains the detected issues and rewrites text in a chosen voice. It is available through Claude Code, Codex, Gemini, DeepSeek, browser and command-line workflows, with a Chrome extension and a LoRA-Cadence training workflow.
Public prototype for transforming rough ideas into structured Veo 3 video-generation prompts, initially focused on fashion films, editorials, runway, beauty and luxury campaigns. It combines prompt retrieval, visual models, generation controls and structured prompt assembly, and is also intended for short films, music videos and visualizers.
Wire-ready data-acquisition architecture simulator for validating packet protocols, codecs, communication links, synchronization, loss/corruption/reordering behavior, record/replay and ground-station tooling before flight hardware is available. It includes Python, C and C++ codec paths and hardware-in-the-loop seams.
Market Position
Wu's portfolio is positioned around small, open-source, practical tools rather than a single commercial platform: MemoryWhale emphasizes local debugging memory instead of hosted agent memory; WireDAQ emphasizes an honest, testable avionics/DAQ harness rather than certified flight software or wire compatibility with MAVLink/CCSDS; Real combines prompt retrieval and visual models for Veo 3 rather than offering a generic text-to-video interface; and Cadence focuses on explainable voice and rhythm editing rather than opaque rewriting. The portfolio's differentiator is the combination of privacy, inspectability, cross-platform integrations and direct fit to specialized robotics/creative workflows.
Founding Story
This is a personal portfolio rather than a separately incorporated venture. Wu says most projects begin with a problem in her own workflow: coding agents forgetting prior debugging knowledge led to MemoryWhale; Veo 3 ignoring parts of creative requests led to Real; and the delay and expense of flight hardware led to WireDAQ, a simulator for testing data-acquisition architecture before hardware integration. Her stated design principle is to make technical tools fast, private, accessible and useful in everyday work.
Business Model
Revenue Model
The available sources describe Wu's work as open-source projects and public prototypes. MemoryWhale is MIT-licensed and Real is described as a public prototype with potential future commercialization; no current paid subscription, API pricing or commercial revenue model is documented.
Target Markets
- Developers and coding-agent users
- Robotics and autonomous-vehicle teams
- Embedded-systems and avionics developers
- AI-video creators, fashion filmmakers and visual storytellers
- Writers and teams seeking voice-aware editing tools
- Recovering previous terminal failures, fixes and debugging context across sessions, machines, worktrees and coding agents.
- Building and testing autonomous underwater vehicle localization, SLAM, state estimation and mission-planning systems.
- Developing embedded sensor, telemetry, DAQ and avionics software before scarce or expensive flight hardware is available.
- Creating fashion films, runway/editorial concepts, luxury campaigns, short films, music videos and visualizers with Veo 3.
- Editing professional, academic or social writing to reduce robotic-sounding phrasing while preserving meaning and personal voice.