v1.0.4 Stable

SIFRA Engine API

The Sifra API allows you to programmatically access the Autonomous Data Scientist Engine. Integrate automated data cleaning, predictive modeling, and insight generation directly into your own applications, dashboards, or ETL pipelines.

RESTful Architecture

Standard HTTP methods and JSON responses for easy integration with any stack.

Enterprise Security

All data is encrypted in transit and at rest. Strict access controls via API keys.

Authentication

Authenticate your API requests by including your API key in the Authorization header of every request. You can manage your API keys in the Console Dashboard.

Header Format

Authorization: Bearer sifra_live_sk_8f92...

Your API keys carry many privileges, so be sure to keep them secure! Do not share your secret API keys in publicly accessible areas such as GitHub, client-side code, and so forth.

Rate Limits

API usage is limited based on your subscription tier. Rate limits are applied per API key.

Tier	Requests / Min	Concurrent Analysis
Starter	60	1
Pro	1,000	10
Enterprise	Unlimited	Unlimited

POST

/v1/datasets/upload

Upload a CSV or JSON dataset to the Sifra Engine. The engine will perform initial profiling and return a dataset ID for further analysis.

curl -X POST https://api.sifra.ai/v1/datasets/upload \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: multipart/form-data" \
  -F "file=@sales_data_q3.csv"

import requests

url = "https://api.sifra.ai/v1/datasets/upload"
headers = {"Authorization": "Bearer YOUR_API_KEY"}
files = {"file": open("sales_data_q3.csv", "rb")}

response = requests.post(url, headers=headers, files=files)
print(response.json())

const axios = require('axios');
const fs = require('fs');
const FormData = require('form-data');

const form = new FormData();
form.append('file', fs.createReadStream('sales_data_q3.csv'));

axios.post('https://api.sifra.ai/v1/datasets/upload', form, {
  headers: {
    ...form.getHeaders(),
    'Authorization': 'Bearer YOUR_API_KEY'
  }
}).then(response => console.log(response.data));

POST

/v1/engine/analyze

Trigger the HDP-FusionNet algorithm on a specific dataset. This endpoint returns comprehensive insights, detected anomalies, and suggested visualizations.

Request Body

dataset_id (Required) The ID returned from the upload endpoint.
mode (Optional) 'fast', 'balanced', or 'deep'. Default: 'balanced'.
target_col (Optional) Specific column to predict or analyze.

curl -X POST https://api.sifra.ai/v1/engine/analyze \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "dataset_id": "ds_88a91b2c",
    "mode": "deep",
    "target_col": "revenue"
  }'

payload = {
    "dataset_id": "ds_88a91b2c",
    "mode": "deep",
    "target_col": "revenue"
}

response = requests.post("https://api.sifra.ai/v1/engine/analyze", 
                         headers=headers, 
                         json=payload)
print(response.json())

Example Response

{
  "status": "success",
  "analysis_id": "an_77f22x90",
  "summary": "Revenue shows a strong positive correlation with Ad_Spend (0.85). Seasonality detected: Monthly.",
  "insights": [
    {
      "type": "correlation",
      "description": "High correlation between User_Age and Purchase_Frequency."
    },
    {
      "type": "anomaly",
      "description": "3 outliers detected in Q4 data points."
    }
  ],
  "suggested_charts": ["heatmap", "time_series_decomposition"]
}

POST

/v1/engine/visualize

Generates a JSON configuration for charts (Plotly/ECharts compatible) or a static image based on the analysis results.

curl -X POST https://api.sifra.ai/v1/engine/visualize \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "analysis_id": "an_77f22x90",
    "chart_type": "heatmap"
  }'

SDKs & Libraries

sifra-python

Official Python wrapper for Sifra Engine.

pip install sifra-ai

sifra-node

Node.js client for backend integration.

npm install sifra-ai

Error Codes

Code	Description
400	Bad Request - Invalid JSON or missing parameters.
401	Unauthorized - Invalid API key.
429	Too Many Requests - Rate limit exceeded.
500	Internal Server Error - The engine encountered an issue processing the data.