# VectorPrism

One canonical identity. Human page: https://www.insightits.com/products/vectorprism.html

| Field | Value |
|-------|-------|
| Slug | `vectorprism` |
| Status | published |
| Family | supporting |
| Infra category | — |
| Record type | product |
| Parent | — |
| JSON | https://www.insightits.com/catalog/vectorprism.json |
| Markdown | https://www.insightits.com/catalog/vectorprism.md |
| GitHub | https://github.com/insightitsGit/VectorPrism |
| PyPI | https://pypi.org/project/vectorprism/0.1.0/ |
| Install | `pip install "vectorprism==0.1.0". Optional: pip install "vectorprism[all]". Apache-2.0, Python ≥ 3.10. Soft CTA RECOVER — mailto:info@insightits.com?subject=RECOVER.` |
| Version | 0.1.0 |

## What it is

Intent-gated 1024d retrieval using positional subspace multiplexing in one tensor — causal / taxonomy channels. Pin vectorprism==0.1.0.

## Problem

Multi-vector RAG cost and ungated retrieval that ignores intent. Teams need channels in one 1024d tensor, not a second product family.

## Alternatives

- **Multi-vector pgvector without intent gates** — No published competitor comparison page. Soft CTA: RECOVER.

## Features

- **Root-cause / incident logs** — “Why did X fail?” — causal / time ODE channel up-weighted on why / cause / reason queries instead of funny cosine neighbors.
- **Taxonomy / ontology search** — Hierarchy distorts in Euclidean space. Hyperbolic taxonomy channel scored with Poincaré distance in Stage 2.
- **Compliance / bitemporal** — Control header packs timestamp and model version for ingest. Stage-1 currently gates on epistemic truth and anchor distance — those two fields are queried; timestamp / model_version are not Stage-1 reject filters yet.

## Architecture

Positional subspace multiplexing (PSM) in one 1024d tensor with causal and taxonomy channels. pgvector multi-vector cost reduction is the job, not a Graph RAG library.

## Use cases

- Intent-gated retrieval over existing pgvector
- Finance / taxonomy channel separation in one embedding

## Benchmarks

No published vendor benchmark for this product. Do not invent metrics.

### Disclosures

- No published vendor benchmark for this product. Do not invent metrics.



## GitHub

https://github.com/insightitsGit/VectorPrism

## PyPI

https://pypi.org/project/vectorprism/0.1.0/

## Documentation

- [Pilot notes](https://github.com/insightitsGit/VectorPrism/blob/main/PILOT.md)
- [Production notes](https://github.com/insightitsGit/VectorPrism/blob/main/PRODUCTION.md)
- [Finance demo report](https://github.com/insightitsGit/VectorPrism/blob/main/demos/finance_demo/TECHNICAL_REPORT.md)
- [VectorPrism Demo page](https://www.insightits.com/products/vectorprism-demo.html)
- [Compose with Manifest / Shield / Eval](https://www.insightits.com/products/vectorprism.html#compose)
- [https://github.com/insightitsGit/VectorPrism](https://github.com/insightitsGit/VectorPrism)
- [https://pypi.org/project/vectorprism/0.1.0/](https://pypi.org/project/vectorprism/0.1.0/)
- [https://insightitsgit.github.io/VectorPrism/](https://insightitsgit.github.io/VectorPrism/)
- [https://www.insightits.com/products/vectorbridge.html](https://www.insightits.com/products/vectorbridge.html)
- [https://www.insightits.com/products/prismrag.html](https://www.insightits.com/products/prismrag.html)
- [pypiProject](https://pypi.org/project/vectorprism/)
- [docker](https://github.com/insightitsGit/VectorPrism/blob/main/DOCKER.md)

## Is not

- VectorBridge (migration)
- PrismRAG (taxonomy graph)
- PrismCortex (agent memory)
- A production guarantee from the adversarial finance pack (vendor harness; production bar = your eval set)

## Related products

- [prismrag](./prismrag.md)
- [vectorbridge](./vectorbridge.md)
- [vectorprism-demo](./vectorprism-demo.md)
- [prismmanifest](./prismmanifest.md)
- [prism-shield](./prism-shield.md)
- [prism-eval](./prism-eval.md)