Tanuki Labs case studies
These public case studies describe products and research systems built by Tanuki Labs. They focus on the problem, the system created, and the engineering idea each project demonstrates. They do not claim invented client results or unverified performance metrics.
CASE 01AI agent governance platform
ActBound
The challenge
AI agents can act quickly, but operators still need approval boundaries, visible evidence, and a reliable way to stop or review consequential actions.
What Tanuki Labs built
ActBound is a governance and runtime-control platform for approvals, blocked actions, live feeds, proof records, and human oversight before automation touches the real world.
What it demonstrates
Tanuki Labs can turn agent safety and operational governance into a usable product interface rather than leaving those concerns as policy documents.
CASE 02Digital business card system
Contact Shuffle
The challenge
People working across multiple companies and projects need distinct public identities, card links, teams, and QR routes without maintaining disconnected static sites.
What Tanuki Labs built
Contact Shuffle is a server-backed digital business card system for creating, editing, organizing, cloning, publishing, and sharing multi-company profiles.
What it demonstrates
Tanuki Labs can design identity tools around modern work patterns while separating private administration from focused public card experiences.
CASE 03Minimum-state intelligence research
VAMANIR
The challenge
Claims about smaller intelligent systems need a reproducible way to compare memory, architecture, evidence, and quality boundaries without hiding assumptions.
What Tanuki Labs built
VAMANIR is an architecture-neutral research system for investigating the smallest evidence-valid physical form an intelligent system can occupy.
What it demonstrates
Tanuki Labs combines research framing, controlled comparison, and software tooling to make a difficult technical question inspectable.
CASE 04Local AI operating shell
PocketFlow
The challenge
Useful personal AI often depends on remote cloud services even when capable Android hardware is already available and could keep data and control closer to the user.
What Tanuki Labs built
PocketFlow is a local-first AI operating shell that turns an ordinary Android phone into a control room for models, automations, files, relays, and daily work.
What it demonstrates
Tanuki Labs explores accessible AI through existing hardware, modular architecture, local execution, and cloud access by choice rather than requirement.
NEXT STEPNeed a system built around your workflow?
Tanuki Labs combines AI software development, custom engineering, automation, and product design to move from a defined problem to a working release.