Rule Inference API

The Rule Inference API provides functions to infer Vadalog rules from a database or data source schema. It generates a linear Vadalog rule for each table or file and a join rule for each table having foreign keys.

Infer Schema

Python SDK
REST API

from prometheux_chain.data.database import Database

# Create a Database object
db = Database(
    database_type="postgresql",
    username="prometheux",
    password="prometheux",
    host="localhost",
    port=5432,
    database_name="prometheux"
)

# Infer Vadalog rules from the database
inferred_rules = px.infer_schema(db, add_bind=True, add_model=False)

# Save the inferred rules to a file
with open("infer-from-postgresql.vada", 'w') as file:
    file.write(inferred_rules)

Function Signature

def infer_schema(database, add_bind=True, add_model=False)

Parameters

Parameter	Type	Required	Description
`database`	Database	Yes	An instance of the `Database` class containing connection details
`add_bind`	bool	No	Whether to add a bind statement in the inferred schema. Defaults to True
`add_model`	bool	No	Whether to add a model annotation statement. Defaults to False

ReturnsReturns inferred Vadalog rule from database or datasource schema as a string.

HTTP Request

POST /api/v1/data/infer-schema

Parameters

Parameter	Type	Required	Description
database	object	Yes	Database connection configuration
addBind	boolean	No	Whether to add a bind statement (default: true)
addModel	boolean	No	Whether to add a model annotation (default: false)

The database object should contain:

Field	Type	Required	Description
databaseType	string	Yes	Database type (e.g., “postgresql”, “mysql”, “neo4j”, “csv”, “excel”)
host	string	Yes	Database host
port	integer	No	Database port
username	string	Yes	Database username
password	string	Yes	Database password
databaseName	string	Yes	Database name
schema	string	No	Database schema
options	object	No	Additional database-specific options

cURL Example

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "postgresql",
      "host": "localhost",
      "port": 5432,
      "username": "prometheux",
      "password": "prometheux",
      "databaseName": "prometheux"
    },
    "addBind": true,
    "addModel": false
  }'

Response

{
  "data": "@bind(\"customers\", \"postgresql\", ...).\ncustomers(Id, Name, Email) <- ...",
  "message": "Schema inferred successfully",
  "status": "success"
}

Database Examples

PostgreSQL

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="postgresql",
    username="prometheux",
    password="prometheux",
    host="localhost",
    port=5432,
    database_name="prometheux"
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "postgresql",
      "host": "localhost",
      "port": 5432,
      "username": "prometheux",
      "password": "prometheux",
      "databaseName": "prometheux"
    },
    "addBind": true
  }'

Neo4j

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="neo4j",
    username="neo4j",
    password="neo4j2",
    host="localhost",
    port=7687,
    database_name="neo4j"
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "neo4j",
      "host": "localhost",
      "port": 7687,
      "username": "neo4j",
      "password": "neo4j2",
      "databaseName": "neo4j"
    },
    "addBind": true
  }'

CSV File from S3

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="csv",
    username="AKIA4xxxx12",
    password="JyxxxxU+",
    host="s3a://prometheux-data",
    port=None,
    database_name="companies.csv",
    options={
        "region": "eu-west-2", 
        "endpoint": "s3.amazonaws.com",
        "credentials.provider": "org.apache.hadoop.fs.s3a.SimpleAWSCredentialsProvider",
        "delimiter": "\t"
    }
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "csv",
      "host": "s3a://prometheux-data",
      "username": "AKIA4xxxx12",
      "password": "JyxxxxU+",
      "databaseName": "companies.csv",
      "options": {
        "region": "eu-west-2",
        "endpoint": "s3.amazonaws.com",
        "credentials.provider": "org.apache.hadoop.fs.s3a.SimpleAWSCredentialsProvider",
        "delimiter": "\t"
      }
    },
    "addBind": true
  }'

Databricks

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="databricks",
    username="token",
    password="dapixxxx",
    host="dbc-xxxx-02fe.cloud.databricks.com",
    port=443,
    database_name="/sql/1.0/warehouses/3283xxxx"
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "databricks",
      "host": "dbc-xxxx-02fe.cloud.databricks.com",
      "port": 443,
      "username": "token",
      "password": "dapixxxx",
      "databaseName": "/sql/1.0/warehouses/3283xxxx"
    },
    "addBind": true
  }'

Databricks with Specific Schema

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

inferred_rules = px.infer_schema(
    Database(
        database_type="databricks",
        username="token",
        password="dapixxxx",
        host="dbc-xxxx-02fe.cloud.databricks.com",
        port=443,
        database_name="/sql/1.0/warehouses/3283xxxx",
        schema="my_catalog.my_schema"
    ),
    add_bind=True,
    add_model=False
)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "databricks",
      "host": "dbc-xxxx-02fe.cloud.databricks.com",
      "port": 443,
      "username": "token",
      "password": "dapixxxx",
      "databaseName": "/sql/1.0/warehouses/3283xxxx",
      "schema": "my_catalog.my_schema"
    },
    "addBind": true,
    "addModel": false
  }'

Snowflake

Instead of using a password, you can use a Programmatic Access Token (PAT) for authentication. This avoids MFA prompts during automated workflows. See the Snowflake data source documentation for details.

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="snowflake",
    username="my_username",
    password="my_password",  # Or use your PAT token
    host="jdbc:snowflake://A77885826xxxx-IV3xxxx.snowflakecomputing.com",
    port=443,
    database_name="my_database",
    schema="my_schema",
    options={"warehouse": "my_warehouse"}
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "snowflake",
      "host": "jdbc:snowflake://A77885826xxxx-IV3xxxx.snowflakecomputing.com",
      "port": 443,
      "username": "my_username",
      "password": "my_password",
      "databaseName": "my_database",
      "schema": "my_schema",
      "options": {
        "warehouse": "my_warehouse"
      }
    },
    "addBind": true
  }'

Excel File

Excel files are treated as a database where sheets are considered as tables.

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="excel",
    username="",
    password="workbookPassword",
    host="path/to/excel_file",
    port=None,
    database_name="excel_file.xlsx",
)

inferred_rules = px.infer_schema(db, add_bind=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "excel",
      "host": "path/to/excel_file",
      "username": "",
      "password": "workbookPassword",
      "databaseName": "excel_file.xlsx"
    },
    "addBind": true
  }'

BigQuery

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"
gcpAccessToken = os.environ["GCP_ACCESS_TOKEN"]

db = Database(
    database_type="bigquery",
    username="",
    password="",
    host="",
    port=None,
    database_name="my_project_id",
    schema="datasetId",
    options={
        "authMode": "gcpAccessToken",
        "gcpAccessToken": gcpAccessToken,
        "parentProject": "my_parent_project_id",
        "billingProjectId": "my_billing_project_id",
        "region": "us-central1"
    }
)

inferred_rules = px.infer_schema(db, add_bind=True, add_model=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "bigquery",
      "host": "",
      "username": "",
      "password": "",
      "databaseName": "my_project_id",
      "schema": "datasetId",
      "options": {
        "authMode": "gcpAccessToken",
        "gcpAccessToken": "YOUR_GCP_ACCESS_TOKEN",
        "parentProject": "my_parent_project_id",
        "billingProjectId": "my_billing_project_id",
        "region": "us-central1"
      }
    },
    "addBind": true,
    "addModel": true
  }'

Text File

Infer concepts and relationships from text content.

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="text",
    username="",
    password="",
    host="path/to/file",
    port=None,
    database_name="document.txt"
)

inferred_rules = px.infer_schema(db, add_bind=True, add_model=False)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "text",
      "host": "path/to/file",
      "username": "",
      "password": "",
      "databaseName": "document.txt"
    },
    "addBind": true,
    "addModel": false
  }'

Binary File (PDF, Images)

Binary files support various formats including PDF, JPG, PNG, and other binary formats.

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="binaryfile",
    username="",
    password="",
    host="path/to/file",
    port=None,
    database_name="document.pdf"
)

inferred_rules = px.infer_schema(db, add_bind=True, add_model=False)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "binaryfile",
      "host": "path/to/file",
      "username": "",
      "password": "",
      "databaseName": "document.pdf"
    },
    "addBind": true,
    "addModel": false
  }'

Business Documents

Structured documents such as ID documents, receipts, tax forms, and mortgage documents. Supported document types include:

Category	Document Types
Financial	check.us, bankStatement.us, payStub.us, creditCard, invoice
ID Documents	idDocument.driverLicense, idDocument.passport, idDocument.nationalIdentityCard, idDocument.residencePermit, idDocument.usSocialSecurityCard
Receipts	receipt.retailMeal, receipt.creditCard, receipt.gas, receipt.parking, receipt.hotel
Tax Documents	tax.us.1040.2023, tax.us.w2, tax.us.w4, tax.us.1095A, tax.us.1098, tax.us.1099
Mortgage Documents	mortgage.us.1003 (URLA), mortgage.us.1004 (URAR), mortgage.us.closingDisclosure
Other	contract, healthInsuranceCard.us, marriageCertificate.us

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="binaryfile",
    username="",
    password="",
    host="path/to/file",
    port=None,
    database_name="driver_license.pdf",
    options={"documentType": "idDocument.driverLicense"}
)

inferred_rules = px.infer_schema(db, add_bind=True, add_model=False)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "binaryfile",
      "host": "path/to/file",
      "username": "",
      "password": "",
      "databaseName": "driver_license.pdf",
      "options": {
        "documentType": "idDocument.driverLicense"
      }
    },
    "addBind": true,
    "addModel": false
  }'

Amazon DynamoDB

Python SDK
REST API

from prometheux_chain.data.database import Database

os.environ["PMTX_TOKEN"] = "YOUR_TOKEN"

db = Database(
    database_type="dynamodb",
    username="AKIA4xxxx12",  # AWS Access Key ID
    password="JyxxxxU+",     # AWS Secret Access Key
    host="",
    port=None,
    database_name="",
    options={
        "region": "us-east-1",
        "endpoint": "",
        "sessionToken": "",
        "sampleLimit": "100"
    }
)

inferred_rules = px.infer_schema(db, add_bind=True, add_model=True)

curl -X POST "https://api.prometheux.ai/jarvispy/my-org/my-user/api/v1/data/infer-schema" \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  -d '{
    "database": {
      "databaseType": "dynamodb",
      "host": "",
      "username": "AKIA4xxxx12",
      "password": "JyxxxxU+",
      "databaseName": "",
      "options": {
        "region": "us-east-1",
        "endpoint": "",
        "sessionToken": "",
        "sampleLimit": "100"
      }
    },
    "addBind": true,
    "addModel": true
  }'

Introduction

Get Started

Platform

Integrations

Examples

Vadalog Reference

Resources

Rule Inference API

Rule Inference API

Infer Schema

HTTP Request

Parameters

cURL Example

Response

Database Examples

PostgreSQL

Neo4j

CSV File from S3

Databricks

Databricks with Specific Schema

Snowflake

Excel File

BigQuery

Text File

Binary File (PDF, Images)

Business Documents

Amazon DynamoDB

Introduction

Get Started

Platform

Integrations

Examples

Vadalog Reference

Resources

​Rule Inference API

​Infer Schema

​HTTP Request

​Parameters

​cURL Example

​Response

​Database Examples

​PostgreSQL

​Neo4j

​CSV File from S3

​Databricks

​Databricks with Specific Schema

​Snowflake

​Excel File

​BigQuery

​Text File

​Binary File (PDF, Images)

​Business Documents

​Amazon DynamoDB

Rule Inference API

Infer Schema

HTTP Request

Parameters

cURL Example

Response

Database Examples

PostgreSQL

Neo4j

CSV File from S3

Databricks

Databricks with Specific Schema

Snowflake

Excel File

BigQuery

Text File

Binary File (PDF, Images)

Business Documents

Amazon DynamoDB