Jobs API

The Jobs API allows you to create and manage scheduled synchronization jobs that automatically sync data from source tables to knowledge bases at regular intervals.

Overview

Jobs provide an automated way to keep your knowledge bases up-to-date by continuously syncing data from your database tables. You can configure filters, specify content columns, and set custom schedules to meet your data synchronization needs.

---

Create a Job

Creates a new scheduled job to sync data from a source table to a knowledge base.

Endpoint

POST /jobs/create

Request Body

The request body should contain a JSON object with the following properties:

Parameter	Type	Required	Description
job_name	string	✅	Unique identifier for the job
source_table	string	✅	Source table in format database.table
kb_name	string	✅	Target knowledge base name
content_columns	array<string>	✅	Columns to include in the sync
filter_column	string	✅	Column used for filtering data
interval	string	✅	Sync frequency (e.g., "10 MINUTES", "1 HOUR")
start_time	string	✅	Job start time in ISO format
end_time	string	✅	Job end time in ISO format

Example Request

{
  "job_name": "sync_orders",
  "source_table": "orders_db.orders",
  "kb_name": "orders_kb",
  "content_columns": ["product", "description"],
  "filter_column": "created_at",
  "interval": "10 MINUTES",
  "start_time": "2025-06-19T18:00:00",
  "end_time": "2025-06-20T18:00:00"
}

Response

Success Response

Status Code: 201 Created

{
  "success": true,
  "message": "Job created successfully",
  "data": {
    "job_id": "job_12345",
    "job_name": "sync_orders",
    "status": "scheduled",
    "created_at": "2025-06-19T12:00:00Z",
    "next_run": "2025-06-19T18:00:00Z"
  }
}

Error Response

Status Code: 400 Bad Request

{
  "success": false,
  "error": {
    "code": "INVALID_REQUEST",
    "message": "Invalid job configuration",
    "details": [
      "job_name is required",
      "invalid interval format"
    ]
  }
}

cURL Example

curl -X POST "http://127.0.0.1:8000/jobs/create" \
  -H "Content-Type: application/json" \
  -d '{
    "job_name": "sync_orders",
    "source_table": "orders_db.orders",
    "kb_name": "orders_kb",
    "content_columns": ["product", "description"],
    "filter_column": "created_at",
    "interval": "10 MINUTES",
    "start_time": "2025-06-19T18:00:00",
    "end_time": "2025-06-20T18:00:00"
  }'

Python Example

import requests
import json

url = "http://127.0.0.1:8000/jobs/create"
payload = {
    "job_name": "sync_orders",
    "source_table": "orders_db.orders",
    "kb_name": "orders_kb",
    "content_columns": ["product", "description"],
    "filter_column": "created_at",
    "interval": "10 MINUTES",
    "start_time": "2025-06-19T18:00:00",
    "end_time": "2025-06-20T18:00:00"
}

response = requests.post(url, json=payload)
print(response.json())

JavaScript Example

const response = await fetch('http://127.0.0.1:8000/jobs/create', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    job_name: 'sync_orders',
    source_table: 'orders_db.orders',
    kb_name: 'orders_kb',
    content_columns: ['product', 'description'],
    filter_column: 'created_at',
    interval: '10 MINUTES',
    start_time: '2025-06-19T18:00:00',
    end_time: '2025-06-20T18:00:00'
  })
});

const data = await response.json();
console.log(data);

---

Interval Formats

The interval parameter accepts various time formats:

Format	Description	Example
X MINUTES	Every X minutes	"5 MINUTES", "30 MINUTES"
X HOURS	Every X hours	"1 HOUR", "6 HOURS"
X DAYS	Every X days	"1 DAY", "7 DAYS"
HOURLY	Every hour	"HOURLY"
DAILY	Every day	"DAILY"
WEEKLY	Every week	"WEEKLY"

---

Error Codes

Code	Description
INVALID_REQUEST	Request body validation failed
DUPLICATE_JOB	Job with the same name already exists
INVALID_TABLE	Source table does not exist or is inaccessible
INVALID_KB	Knowledge base does not exist
INVALID_INTERVAL	Interval format is not supported
INVALID_DATETIME	Start or end time format is invalid

---

Notes

• Job names must be unique across the system
• The start_time and end_time should be in ISO 8601 format
• Jobs will automatically stop running after the end_time
• The filter_column is typically a timestamp column used to identify new or updated records
• All specified content_columns must exist in the source table

---

Rate Limits

• Maximum of 100 active jobs per account
• Minimum interval is 1 minute
• Maximum job duration is 30 days