Web App for SQL Database Interaction _ Supports RAG with Timescale and PGVector _ Interact with Ollama via widget _ Upload query-response to Pinata

Editor

A semi-colon is automatically added to the query if absent before it is executed on the database
Toggle off connection pooling if the database query takes too long and try again. This development will not hold for local databases as well as remote services that support pooling through the URL like Neon.tech

Use `toSql` in the format below to to convert or pad vector embeddings

toSql(embedding, vector_type i.e vector(float8), vector(int4) and vector(text), normalization options - l2_normalize or cosine_normalize, padding_value - eg 0, padding_length - 128, 256, 1536...)

Ctrl + ←↑ ↓→ quickens the cursor movements by making it jump over words.
Shift + the key combination above highlights words at once.
page.result, page.embedding, page.metadata are available for viewing or modification via the Console

Run the query below to get all available extensions

SELECT * FROM pg_available_extensions WHERE name = 'vector';

Retrieval Augmented Generation

Requires a database that support vector embeddings such as Timescale ↗

Create Schema

Click on each query below to edit them before applying them to the database

Embedding from Ollama
The requested embedding is automatically stored and used in this application if successful.Click on the button below to open the Ollama widget
Add seed data to database
Click on each to expand/edit in editor

database

server address

server port

Execute Queries (the templates are editable)

SELECT
  table_cell_id, 
  first_field
FROM
  table_name
WHERE
  table_cell_id = 1;