IP Vault

​

What is the IP Vault?

The IP Vault is your personal file storage. When you upload documents here, your digital twin can read and reference them when answering your questions. Simple workflow:

1. Upload your files (PDFs, documents, spreadsheets, etc.)
2. Ask questions about your content
3. Get answers that reference your uploaded materials

​

When Does My Digital Twin Use My Files?

Your digital twin automatically searches your uploaded files when you:

• Ask questions related to content you’ve uploaded
• Select specific files to include in a conversation
• Reference topics covered in your documents

---

​

Two retrieval modes — RAG and KAG

Every file goes through two complementary pipelines after upload: Both pipelines run automatically; you don’t have to choose between them. When KAG is enabled — by your institution for its shared vault, or by you for your personal files via the “Build a knowledge graph from my files” option in your profile — your Digital Twin can combine vector recall with graph traversal so answers are grounded in both the direct text and the structured relationships derived from it. You can also browse the resulting graph in the Tree of Knowledge view.

---

​

KAG Support: meaning-aware retrieval

Plain RAG is excellent at finding the right passage, but some questions need the right connections — how people, policies, and concepts relate across many documents. That’s what KAG (Knowledge-Augmented Generation) adds. When KAG Fusion is enabled for a vault, Pria builds a knowledge graph — a “tree of knowledge” — from your documents, extracting the entities and the relationships between them, and then fuses that graph with the dense vector search so a single answer draws on both. KAG always sits on top of RAG; it augments retrieval, it never replaces it.

​

How a grounded answer comes together

1. Content sources — documents, course files, policies, videos, and data tables become searchable once indexed.
2. RAG finds relevant passages — semantic similarity surfaces the specific text that matches your question (precise lookup).
3. KAG extracts meaning — entities (people, topics, dates) and the relationships between them form a living knowledge map, connecting ideas that no single passage spells out.
4. Better conversations — answers combine the direct text with that structured context, and every claim traces back to a source file you can open.

​

KAG vs GraphRAG

A popular alternative pattern, GraphRAG, answers from a knowledge graph alone. Searching only the graph tends to miss passages that were never distilled into an entity, leans on higher token usage to walk the graph, and is harder to trace back to an original source. Pria’s KAG keeps the dense RAG leg and adds the graph plus an audit trail of evidence, so you get the best of both: The payoff shows up most on complex, multi-hop questions (“how does this policy affect that team’s approvals?”) where the answer threads through several documents rather than sitting in one paragraph.

---

​

Relevance Scores

When your digital twin finds matching content in your files, it shows a relevance score indicating how semantically close the match is to your question. Matches are found by meaning, not just keywords, and scored on a 0–100% scale — the closer a passage is to what you asked, the higher its score. By default, results below 45% relevance are filtered out. Administrators can adjust this minimum relevance threshold to control the trade-off between recall (more results) and precision (more relevant results).

---

Technical Details: How File Search Works

Your digital twin uses a technology called RAG (Retrieval Augmented Generation) to search your files. Here’s what happens behind the scenes:

1

File Processing

When you upload a file, the system breaks it into smaller sections (segments) and converts each section into a numerical representation called a vector embedding — a high-dimensional array of numbers that captures the semantic meaning of the text. Embeddings are generated by a dedicated AI model (e.g., OpenAI text-embedding-3-small, Amazon titan-embed-text-v2, or Google gemini-embedding-001).

2

Question Analysis

Your question is converted into a vector embedding using the same model, allowing the system to mathematically compare your question against all your file sections in vector space.

3

Finding Matches

Pria compares the meaning of your question against every indexed section of your files and ranks the closest matches by relevance score.

4

Relevance Threshold

Results below a minimum relevance threshold are filtered out (default: 45%). Administrators can adjust how strict the matching is:

• Higher threshold: Only highly relevant content is included — better precision
• Lower threshold: More content is included — better recall, but may be less directly relevant

​

Segment Size and Search Accuracy

When a file is processed, it is split into segments (also called “chunks”) before being converted into vector embeddings. The size of these segments has a direct impact on how accurately your digital twin retrieves relevant content.Research findings from studies by Chroma Research and multi-dataset analyses consistently show an inverse relationship between segment size and retrieval precision:

Segment Size	Precision	Recall	Trade-off
Small (~200 tokens)	Highest	Good	Best precision, but may fragment context
Medium (~400 tokens)	Moderate	Highest	Balanced precision and recall
Large (~800 tokens)	Lowest	Good	More context per segment, but more noise in results

Key insights from the research:

• Smaller segments match queries more precisely because the vector embedding represents a focused piece of content rather than a broad summary of many topics.
• Larger segments preserve more surrounding context but introduce noise — the embedding averages over more content, diluting the signal for any single topic.
• Industry defaults (e.g., 800 tokens with 400-token overlap) were found to score below average across all retrieval metrics.
• The optimal size depends on query type: fact-based questions benefit from smaller segments (256-512 tokens), while analytical questions requiring broader context may benefit from larger ones (1024+ tokens).

At Praxis, we have determined that segment length should not exceed 8,000 characters, limiting each segment to approximately 2,000 tokens maximum. This cap accommodates the input limits of all supported embedding models — including OpenAI, Amazon Titan, and Google Gemini — while keeping segments small enough to preserve semantic search accuracy. This balance ensures that retrieval precision remains high without fragmenting content so aggressively that context is lost.

You can view and edit your file’s segments using the RAG Segments action in your file management options. Each segment shows its character count and the embedding model used to index it.