📦 RAG Vector Base Writer v7 🚀

AI Knowledge Base
Management System

Comprehensive system for working with vector databases. Upload documents, process content with AI, vectorize and search semantically. Full integration with self-learning system.

3
API Endpoints
4
Operation Modes
6
File Types
5
Vector DBs
768
Vector Size
📄 Multi-format Upload
🤖 AI Fragmentation
🧬 Gemini Vectorization
🔄 Self-Learning Stats

😤 Knowledge Management Pain Points

Why manual knowledge base population is painful

📁

Scattered Formats

Documents in PDF, Word, Excel, presentations... Each format needs its own parser. Collecting everything together takes hours of manual work.

✂️

Poor Fragmentation

Is splitting text into chunks simple? No. Breaking in the middle of a thought kills semantics. Chunks too large mean poor search.

🔍

Bad Semantic Search

Without quality embeddings and proper indexes, searching the knowledge base becomes a lottery.

☁️

Files Are Scattered

Some documents in Google Drive, some in Dropbox, something in OneDrive. Gathering everything in one place is a quest.

🔄

No Versioning

Updated a document — how to update the database entry? Delete old, add new, edit existing?

📊

No Statistics

How many records in the database? Which tables are updated most actively? How is self-learning working? Complete unknown.

✨ Key Features

Everything you need to manage AI agent knowledge

📥

Multi-format Upload

PDF, DOCX, XLSX, CSV, TXT, PPTX — all popular document formats. Upload files directly or via URL from cloud storage.

🤖

AI Content Processing

Intelligent fragmentation with Gemini AI. Automatic metadata generation: categories, keywords, topics.

🧬

768-dimensional Vectorization

Embeddings via Google Gemini Embeddings. HNSW indexes in Supabase for lightning-fast semantic search.

🗄️

Dual Storage

Structured tables in PostgreSQL + vectorization in Supabase. Maximum flexibility for any tasks.

🔒

Security

API key authorization, rate limiting (30 req/min), table and mode validation. Protection from unauthorized access.

📊

Self-Learning Stats

Detailed self-learning statistics: cycles, dialogues, approved/rejected updates, activity by tables.

🔄 4 Operation Modes

Full control over data in vector databases

APPEND

Add new records to the end of the table without deleting existing ones. Ideal for gradually building up the knowledge base.

Safe addition
🔄

REPLACE

Complete table cleanup and replacement of all data with new content. For full knowledge base rewrite.

Full rewrite
✏️

EDIT

Update a specific record by ID. For pinpoint changes to individual records without affecting others.

Requires editId
🗑️

DELETE

Delete record by ID with cascading deletion from related tables (document_rows, document_metadata).

Cascading delete

📄 6 File Types

Support for all popular document formats

📄

PDF

Extract PDF Text node

📝

TXT

Direct text extraction

📘

DOCX

ZIP → XML → Parse

📊

XLSX/XLS

Dual storage mode

📋

CSV

Table + Vector

📑

PPTX

Extract slides text

☁️ Cloud Sources

Automatic URL recognition and conversion

📝

Google Docs

docs.google.com

→ DOCX export
📊

Google Sheets

spreadsheets

→ XLSX export
📑

Google Slides

presentation

→ PPTX export
☁️

Google Drive

drive.google.com

→ Direct download
📦

Dropbox

dropbox.com

→ dl=1 parameter
💾

OneDrive

onedrive.live.com

→ Direct support

🤖 AI Content Processing

Intelligent fragmentation and metadata generation

🧠

Processing Model

Model Gemini 3 Flash
Temperature 0.2
Embeddings gemini-embedding-001
Chunk Size 1500 characters
Dimensions 768
📏

Fragmentation Rules

Fragment Size 200-1000 chars
List Grouping ✅ Enabled
Context Preservation ✅ Enabled
[FULL] Command Keep whole
🏷️

Auto Metadata

category Content category
keywords Keywords
topic Fragment topic
technique Technique/method
stage Funnel stage

🗄️ 5 Vector Databases

Supabase + HNSW indexes for semantic search

📚

knowledge_base

Product knowledge base — prices, features, algorithms

📜

interaction_policies

Interaction policies — channels, rules, limitations

💼

sales_strategies

Sales strategies — argumentation, objections, closing

📱

contact_capture

Contact capture methods — progressive disclosure

🎭

conversation_scenarios

Dialogue scenarios — use cases, escalation, branching

🔗 Self-Learning Integration

Full synchronization with AI agent self-learning system

🧠
AI Self-Learning
📦
Vector Base Writer
🗄️
Supabase Vector
🤖
AI Core Agent

Closed Improvement Loop

The self-learning system analyzes dialogues, generates prompt improvement suggestions, validates them, and applies them to vector databases through Vector Base Writer. AI Core Agent immediately starts using updated knowledge. The /get-learning-stats endpoint shows detailed statistics of the entire process.

💎 System Advantages

Why Vector Base Writer is the right choice

Fast Upload

One API call to upload a document of any format. The system will detect the type and process it automatically.

🧠

Smart Fragmentation

AI preserves semantic integrity of text. Related thoughts are not broken in the middle of a sentence.

🔍

Accurate Search

768-dimensional embeddings and HNSW indexes provide lightning-fast semantic search.

🔄

Flexible Management

4 operation modes — append, replace, edit, delete. Full control over data.

📊

Transparent Statistics

Stats endpoint shows everything: learning cycles, updates, activity by tables.

🔒

Reliable Protection

API keys, rate limiting, validation. Your data is securely protected.

📋 Technical Specifications

Key system parameters

3
API Endpoints
4
Modes
6
File Types
5
Vector DBs
768
Dimensions
30/m
Rate Limit