Start with the environment.
Navigate recorded zones, inspect saved imagery, and open the objects associated with each area.
PHYSICAL CONTEXT / PERSISTENT MEMORY
The world changes.
Keep the context.
A spatial memory platform for people and AI. Connect environments, objects, observations, and time through evidence you can inspect.
Early product · private workspaces · web first
ENVIRONMENT / 01● RECORD ACTIVE
OPERATIONS BAY
C-04
Equipment identity
Equipment identity
OBS / 002
Recorded evidence
Recorded evidence
T + 19 DAYS
Change over time
Change over time
PLACE → ZONE → OBJECT → OBSERVATION
Conceptual record structure · illustrative data
BUILT AROUND ENVIRONMENTSFacilitiesField sitesLaboratoriesInfrastructureCampuses
ONE CONTINUOUS WORKSPACE
Explore. Capture. Timeline. Ask.
Ground the record.
Attach evidence, review AI photo suggestions, and confirm object identity and location notes.
Preserve what changed.
Add observations and corrections without replacing the original record. Compare evidence across visits.
Follow the evidence.
Query saved environment records and inspect the observations supporting the answer.
THE PLATFORM DIRECTION
Memory that outlasts a visit.
The goal is a persistent representation of physical environments that people and AI can use across visits and workflows. Object identity, location, provenance, and change belong in the same record.
THE PRODUCT TODAY
Evidence before inference.
Private accounts, environment and zone organization, visual evidence, object histories, AI-assisted photo suggestions, comparisons, and grounded questions are available now.
Cross-visit automatic recognition, measured spatial coordinates, live 3D mapping, shared operational workflows, and a public API remain future work.
THE NEXT PROOF
Return. Recognize. Explain.
Our next validation target: revisit a real environment, reconnect objects to their histories, and explain changes with evidence. We will measure matching accuracy, correction effort, and time saved. These are validation goals, not published performance claims.
Designed for real context.
Is this a home inventory product?
No. The data model is environment → zone → object → observation. A home is one possible environment; the platform direction includes operational and institutional spaces.
What can I try without signing in?
The interactive operations-bay prototype shows five fictional objects, their observation histories, comparisons, and written example answers. It is explicitly labeled and makes no live AI requests.
How does AI use the evidence?
When requested, photo analysis sends the selected image and zone context to OpenAI. Questions use your saved records. You review suggestions before they become saved memories.
Are environments private?
Records and evidence are scoped to your account. The current admin dashboard manages account access rather than browsing members’ records. Team sharing is not yet available.
START WITH ONE ENVIRONMENT
Build the record. Test the value.
Capture a zone, document five objects, and return with new evidence.
Create a workspace ↗