Awareness-AI · Qdrant Manager

Qdrant Manager

Vector Database Management

Connecting...

Enhanced Data Ingestion

Upload files or structured JSON with advanced chunking, legal structure extraction, and quality validation.

Ingestion Mode

Target Collection

Select Files

Vector Options

Generate Dense Vectors Generate Sparse Vectors

Target Collection

Select Legal Documents Supported: PDF, TXT, MD. Metadata files (.meta.json) will be loaded automatically if present.

Document Metadata

Framework Type

Jurisdiction

Legal Domain

Specificity Level

Language

Issuing Authority (Optional)

Chunking Configuration

Chunk Size (chars) Recommended: 1500-2000

Overlap (chars) Preserves context

Vector Types

Dense Sparse

How it works:
• Text is split at sentence boundaries for natural chunks
• Overlap ensures context continuity between chunks
• Legal structure (articles, sections) automatically extracted
• Each chunk preserves position metadata for reconstruction

Quality Validation

Validate before ingestion Skip existing documents Extract legal structure Calculate specificity scores

Validation checks:
• Minimum chunk length (50 chars)
• Maximum chunk length (5000 chars)
• Required metadata presence
• Specificity score calculation (0-1)

Target Collection

Data Type

Select JSON Files Multiple JSON files can be selected. Each file can contain a single object or an array of objects.

Vector Collections

Search with Text

Find relevant documents in a collection using a text query.

Collection to Search

Search Query

Number of Results

Chat with Data (RAG)

Ask questions about your data. The AI will find relevant context and provide an answer.

Select a Collection to Chat With

Chat Configuration

AI Assistant (Optional)

Model (if no assistant selected)

API Endpoint

Context Search Limit

Include search results in context Include collection info Stream responses

Hello! I'm your AI assistant for exploring the data in your Qdrant collections. Select a collection and ask me anything about its content.

Request Payload

{"message": "Select a collection and ask a question to see the payload"}

Step Builder

Create and save individual analytical steps that can be reused across workflows.

Step Name

Step Type

Data Source (Collection)

Step Description / Goal

Workflow Builder

Select and arrange saved steps to build your workflow.

Load Saved Workflow

1. Select Steps

Click to select saved steps...

2. Arrange and Execute

No steps selected. Choose steps from the list above.

Enhanced Data Ingestion

Document Metadata

Chunking Configuration

Quality Validation

Selected Files:

Vector Collections

Create New Collection

Search with Text

Chat with Data (RAG)

Chat Configuration

Request Payload

Step Builder

Workflow Builder

Enhanced Data Ingestion

Document Metadata

Chunking Configuration

Quality Validation

Selected Files:

Vector Collections

Create New Collection

Search with Text

Chat with Data (RAG)

Chat Configuration

Request Payload

Step Builder

Workflow Builder

Collection Details

Find Connections Between Collections

Create Legal/Compliance Collection

1 Collection Information

Vector Configuration