1. The creative problem
Studios produce thousands of outputs. They need immutable revisions, typed dependencies, and full provenance to ship with confidence.
What slows creative teams down is not making new work. It is losing context: which version is correct, how it was made, and what depends on it. We built Kumiho to solve that with a graph-native system for revisions and relationships, while files stay on your own storage.
Then we noticed AI agents have the same structural problem. They lose context across sessions, cannot trace why a decision was made, and drift when memory is implicit.
Kumiho applies one architecture to both worlds: versioned creative assets and persistent AI cognitive memory with provenance and auditability.
We are a small team of creators and pipeline engineers. We have spent years inside VFX, animation, and AI workflows where high-value work gets buried in folder chaos.
Kumiho started as our attempt to make revision tracking and dependency intelligence available without the infrastructure burden. It has grown into a graph-native platform used by creative teams and AI builders.
Not two products. One graph applied twice.
Studios produce thousands of outputs. They need immutable revisions, typed dependencies, and full provenance to ship with confidence.
Agents also produce evolving outputs: decisions, preferences, and facts. Without structure, memory becomes shallow recall with weak auditability.
The same graph primitives solve both. Revisions preserve history, typed edges preserve meaning, and retrieval keeps context operational.
| Primitive | Creative assets | AI memory |
|---|---|---|
| Immutable revisions | Version history for renders/models | Belief history for agent decisions |
| Typed dependencies | Impact analysis for upstream changes | Evidence chains and supersession links |
| BYO storage | Raw files stay on local/NAS/S3 | Conversation artifacts stay local |
| Hybrid retrieval | Find by metadata, text, or relation | Recall by meaning, path, or exact text |
Three shifts that made this inevitable
Creative generation and agent interactions both produce variants faster than humans can manually track.
Studios and AI teams now need reproducibility, explainability, and audit trails as baseline requirements.
Reliable triggers require explicit relationships. Graph context enables safe automation at scale.
When work becomes automated, your knowledge system has to become operational.
Raw content stays in your infrastructure. Kumiho stores metadata, relationships, and pointers.
Revisions are immutable. Updates happen through new revisions, never hidden overwrites.
Vector similarity helps discovery, but typed graph edges encode why things are connected.
Switch tools or LLM providers without losing accumulated context and provenance.
Consolidation uses guardrails: no silent mass deprecation, no unsafe mutation of published records.
Shipping software, not slideware
Browse revisions, compare versions, and trace lineage from one interface.
Capture prompts, seeds, and generation metadata as structured revision history.
Trigger downstream workflows when assets or memory records change.
Give MCP-compatible agents persistent memory with audit-ready provenance.
Build custom workflows and integrations on top of the graph primitives.
Join creators, studios, and AI builders who use Kumiho to track what matters.
Free tier available. No forced uploads. No lock-in.