VectorShift

https://docs.vectorshift.ai
Admin
VectorShift is a platform for building generative AI applications and workflow automations quickly,...
Tokens:116,628
Snippets:1,673
Trust Score:8.8
Update:5 months ago
Show doc for...
Tokens:
Raw
### Text to SQL Node Example Pipeline

Source: https://docs.vectorshift.ai/platform/pipelines/logic/text-to-sql

This example demonstrates a pipeline using the Text to SQL node. It includes a Text node for the natural language query, the Text to SQL node itself, and an Output node to display the generated SQL query.

```text
1. Text Node: Stores the natural language query
   * Text: `Summarize the data`
2. Text to SQL Node: Generates the SQL query
   * SQL Database: MySQL
   * Natural language query: `{{input_0.text}}`
   * Schema: `CREATE TABLE Person (
      id INT AUTO_INCREMENT PRIMARY KEY,
      name VARCHAR(100) NOT NULL,
      phone_number VARCHAR(20),
      email VARCHAR(100) UNIQUE,
      created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP
     );`
3. Output Node: Outputs the SQL query
   * Output: `{{nl_to_sql_0.sql_query}}`
```

--------------------------------

### Start and Continue Conversation in Python

Source: https://docs.vectorshift.ai/platform/interfaces/chatbots/api

Demonstrates how to initiate a new conversation by omitting the conversation_id and then continue an existing one by including it in subsequent requests.

```python
import requests

API_KEY = "your_api_key"
CHATBOT_ID = "your_chatbot_id"

url = "https://api.vectorshift.ai/api/chatbots/run"

headers = {
    "Content-Type": "application/json",
    "Api-Key": API_KEY,
}

# First message (no conversation_id)
payload = {
    "input": "Where is my order #12345?",
    "chatbot_id": CHATBOT_ID,
}

response = requests.post(url, json=payload, headers=headers)
data = response.json()

print(data["output"])
conversation_id = data["conversation_id"]

# Follow-up message (with conversation_id)
payload = {
    "input": "Can I change the shipping address?",
    "chatbot_id": CHATBOT_ID,
    "conversation_id": conversation_id,
}

response = requests.post(url, json=payload, headers=headers)
data = response.json()

print(data["output"])
```

--------------------------------

### Sentence Segmentation Example

Source: https://docs.vectorshift.ai/platform/pipelines/knowledge/text-chunking

Illustrates splitting text into sentences using punctuation and grammatical rules. This method is effective for preserving complete thoughts.

```text
Text splitting is useful.
It helps AI process information.
```

--------------------------------

### Paragraph Segmentation Example

Source: https://docs.vectorshift.ai/platform/pipelines/knowledge/text-chunking

Shows how to segment text into paragraphs, where blocks of text separated by blank lines are treated as individual units. This preserves topic coherence within paragraphs.

```text
Text splitting is useful.
It helps AI process information.

Another key concept is chunk overlap.
```

--------------------------------

### Start and Continue Conversation in JavaScript

Source: https://docs.vectorshift.ai/platform/interfaces/chatbots/api

Shows how to start a new conversation without a conversation ID and then continue it by passing the conversation ID from the initial response.

```javascript
const API_KEY = "your_api_key";
const CHATBOT_ID = "your_chatbot_id";

const url = "https://api.vectorshift.ai/api/chatbots/run";

// First message (no conversation_id)
const firstResponse = await fetch(url, {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Api-Key": API_KEY,
  },
  body: JSON.stringify({
    input: "Where is my order #12345?",
    chatbot_id: CHATBOT_ID,
  }),
});

const firstData = await firstResponse.json();
console.log(firstData.output);

const conversationId = firstData.conversation_id;

// Follow-up message (with conversation_id)
const followUpResponse = await fetch(url, {
  method: "POST",
  headers: {
    "Content-Type": "application/json",
    "Api-Key": API_KEY,
  },
  body: JSON.stringify({
    input: "Can I change the shipping address?",
    chatbot_id: CHATBOT_ID,
    conversation_id: conversationId,
  }),
});

const followUpData = await followUpResponse.json();
console.log(followUpData.output);
```

--------------------------------

### Create Knowledge Base

Source: https://docs.vectorshift.ai/api-reference/knowledge-bases/create

Creates a new knowledge base with specified configurations. Authentication is required.

```APIDOC
## POST /knowledge-base

### Description
Creates a new knowledge base.

### Method
POST

### Endpoint
/knowledge-base

### Request Body
- **name** (string) - Required - Name of the knowledge base
- **description** (string) - Optional - Description of the knowledge base
- **file_processing_implementation** (string) - Optional - Specifies the file processing method. Enum: Default, Textract, Unstructured, LlamaParse, Other.
- **chunk_size** (integer) - Optional - Size of document chunks.
- **chunk_overlap** (integer) - Optional - Overlap between document chunks.
- **analyze_documents** (boolean) - Optional - Whether to analyze documents.
- **apify_key** (string) - Optional - Apify API key.

### Request Example
```json
{
  "name": "My New Knowledge Base",
  "description": "This is a test knowledge base.",
  "file_processing_implementation": "Default",
  "chunk_size": 1000,
  "chunk_overlap": 200,
  "analyze_documents": true,
  "apify_key": "your-apify-key"
}
```

### Response
#### Success Response (200)
- **status** (string) - Enum: success, failed - Status of the operation.
- **id** (string) - ID of the created knowledge base.

#### Response Example
```json
{
  "status": "success",
  "id": "kb_12345abcde"
}
```

#### Error Response (400)
- **status** (string) - Enum: failed - Status of the operation.
- **error** (string) - Description of the error.

#### Error Response Example
```json
{
  "status": "failed",
  "error": "Invalid input provided."
}
```
```

--------------------------------

### Cron Trigger Node Example: Send Email Every Minute

Source: https://docs.vectorshift.ai/platform/pipelines/start/trigger/cron-trigger

This example demonstrates a pipeline that sends an email every minute. It uses a Cron Trigger Node with a cron expression set to '* * * * *' to achieve minute-level scheduling. The pipeline also includes Text, LLM, and Gmail nodes to construct and send the email.

```pipeline
1. Cron Trigger Node: The cron trigger node
   * Cron Expression: `* * * * *`
   * Enable Automation: `true`
2. Text Node: Question to answer with the current timestamp
   * Text: `What is the tallest builiding in NY? {{trigger_0.timestamp}}`
3. LLM Node: Uses the LLM to answer the query and write the email body
   * System (Instruction): `Write an email answers the question based on Context. Only write the email body, not the subject`
   * Prompt: `Question: {{text_0.text}}`
4. Gmail Node: Sends the generated email
   * Action: `Send Email`
   * To: `sample@vectorshift.ai`
   * Subject: `{{trigger_0.timestamp}}`
   * Body: `{{openai_0.response}}`
```

--------------------------------

### Run Knowledge Base Search (HTTP)

Source: https://docs.vectorshift.ai/platform/knowledge/api

This is a raw HTTP request example for calling the knowledge base search endpoint. Ensure you replace 'YOUR_API_TOKEN' with your Bearer JWT token and '{id}' with your knowledge base ID.

```http
POST /search/{id}/run HTTP/1.1
Host: api.vectorshift.ai
Authorization: Bearer YOUR_API_TOKEN
Content-Type: application/json

{
  "input_message": "Sample input message here",
  "conversation_id": "text",
  "is_deployed": true,
  "alpha": 0.5
}
```

--------------------------------

### Configure LLM System Prompt

Source: https://docs.vectorshift.ai/quickstart

Provide instructions to the LLM on how to respond, such as answering questions based on provided context.

```text
Answer the Question based on Context
```

--------------------------------

### Non-Streaming Chatbot Response

Source: https://docs.vectorshift.ai/platform/interfaces/chatbots/api

Example of a successful JSON response from the chatbot API when not using streaming.

```json
{
  "status": "success",
  "output": "Your order #12345 is currently in transit and expected to arrive by Friday.",
  "conversation_id": "abc123def456"
}
```

--------------------------------

### Configure LLM Prompt with Context

Source: https://docs.vectorshift.ai/quickstart

Pass the user's question and relevant data from the knowledge base to the LLM. Labeling the data as 'Question' and 'Context' helps the LLM understand its role.

```text
Question: {{input_0.text}}
Context: {{knowledge_base_0.chunks}}
```

--------------------------------

### Run Chatbot

Source: https://docs.vectorshift.ai/platform/interfaces/chatbots/api

Send a message to a chatbot to get a response. Include a `conversation_id` to continue an existing conversation.

```APIDOC
## POST /api/chatbots/run

### Description
Sends a message to a specified chatbot and returns a response. If `conversation_id` is provided, the conversation continues from the previous state. If `stream` is set to `true`, the response will be delivered as Server-Sent Events.

### Method
POST

### Endpoint
/api/chatbots/run

### Request Body
- **input** (string) - Required - The user's message to the chatbot.
- **chatbot_id** (string) - Required - The ID of the chatbot to interact with.
- **conversation_id** (string) - Optional - The ID of the conversation to continue.
- **stream** (boolean) - Optional - If true, the response will be streamed as Server-Sent Events.

### Request Example
```json
{
  "input": "Where is my order #12345?",
  "chatbot_id": "your_chatbot_id",
  "conversation_id": "abc123def456"
}
```

### Response
#### Success Response (200)
- **output** (string) - The chatbot's response.
- **conversation_id** (string) - The ID of the conversation.

#### Response Example
```json
{
  "output": "Your order #12345 is currently in transit and expected to arrive by Friday.",
  "conversation_id": "abc123def456"
}
```

### Streaming Response Example (Server-Sent Events)
```json
{"output": "Your order", "conversation_id": "abc123def456"}
{"output": "Your order #12345 is currently", "conversation_id": "abc123def456"}
{"output": "Your order #12345 is currently in transit and expected to arrive by Friday.", "conversation_id": "abc123def456"}
```
```

--------------------------------

### Word Segmentation Example

Source: https://docs.vectorshift.ai/platform/pipelines/knowledge/text-chunking

Demonstrates splitting text into individual words. This is a basic method that may lose sentence meaning.

```text
Text splitting is useful.
```
Tokens:
Raw