teradataml 17.20.0.7__py3-none-any.whl → 20.0.0.0__py3-none-any.whl

teradataml/gen_ai/convAI.py

@@ -1,473 +1,473 @@
-# ##################################################################
-#
-# Copyright 2023 Teradata. All rights reserved.
-# TERADATA CONFIDENTIAL AND TRADE SECRET
-#
-# Primary Owner: Kesavaragavan B (kesavaragavan.b@teradata.com)
-# Secondary Owner: Prafulla V Tekawade (prafulla.tekawade@teradata.com)
-#
-# This file implements tdmlAI and DBChain which is used for Vantage 
-# database interaction. The tdmlAI manages LLM inference endpoints and
-# DBChain enables interaction between user, Vantage database.
-# Notes: 
-#   * This code is only for internal use. 
-#   * The code may perform modify, create, or delete operations 
-#     in database based on given query. Hence, limit the permissions 
-#     granted to the credentials.
-# 
-# ##################################################################
-
-# Import required packages.
-import openai
-import os
-from langchain.llms import AzureOpenAI
-from langchain.utilities import SQLDatabase
-from langchain_experimental.sql import SQLDatabaseChain
-from langchain.prompts.prompt import PromptTemplate
-from teradataml import get_context, get_connection
-from teradataml.utils.validators import _Validators
-
-def _set_openAI(api_key, api_type, api_base, api_version):
-    """
-        DESCRIPTION:
-            Internal function to set environment variables for AzureAI.
-            
-        PARAMETERS:
-            api_key:
-                Required Argument.
-                Specifies the AzureAI API key.
-                Types: str
-
-            api_type:
-                Required Argument.
-                Specifies the AzureAI API type.
-                Types: str
-            
-            api_base:
-                Required Argument.
-                Specifies the AzureAI API base url.
-                Types: str
-
-            api_version:
-                Required Argument.
-                Specifies the AzureAI API version.
-                Types: str
-
-        RETURNS:
-            None
-
-        RAISES:
-           None
-
-        EXAMPLES:
-            _set_openAI(api_type = "azure"
-                        api_base = "https://***.openai.azure.com/"
-                        api_version = "2021-12-35"
-                        api_key = "999***")
-        """
-    # Set API type.
-    os.environ["OPENAI_API_TYPE"] = api_type
-    # Set API version like follow "2022-06-10".
-    os.environ["OPENAI_API_VERSION"] = api_version
-    # Set API Base URL as follow 
-    # "https://****instance.openai.azure.com/".
-    os.environ["OPENAI_API_BASE"] = api_base
-    # Set API key.
-    os.environ["OPENAI_API_KEY"] = api_key     
-
-
-class tdmlAI:
-    """tdmlAI provides convenient access to the LLM endpoints for inference."""
-
-    def __init__(self,api_key, 
-                 api_type,
-                 api_base, 
-                 api_version, 
-                 engine,
-                 model_name):
-        """
-        DESCRIPTION:
-            Constructor of tdmlAI that sets up the environment and 
-            initializes the LLM endpoint.
-            
-        PARAMETERS:
-            api_key:
-                Required Argument.
-                Specifies the LLM API key.
-                Types: str
-
-            api_type:
-                Required Argument.
-                Specifies the LLM API type.
-                Types: str
-            
-            api_base:
-                Required Argument.
-                Specifies the LLM API base url.
-                Types: str
-
-            api_version:
-                Required Argument.
-                Specifies the LLM API version.
-                Types: str                       
-                  
-            engine:
-                Required Argument.
-                Specifies the deployment name of the LLM.
-                Types: str           
-
-            model_name:
-                Required Argument.
-                Specifies the LLM model name.
-                Types: str
-
-        RETURNS:
-            None
-
-        RAISES:
-            TeradataMlException, ValueError, TypeError
-
-        EXAMPLES:
-            # Import the modules.
-            from teradataml.gen_ai.convAI import tdmlAI
-            # Create LLM endpoint.
-            tdml_ai_obj = tdmlAI(api_type = "azure",
-                     api_base = "https://****.openai.azure.com/",
-                     api_version = "2000-11-35",
-                     api_key = <provide your llm API key>,
-                     engine = <provide your llm engine name>,
-                     model_name = "gpt-3.5-turbo")
-        """
-
-        # Argument validations
-        awu_matrix = []
-        awu_matrix.append(["api_key", api_key, False, (str)])
-        awu_matrix.append(["api_type", api_type, False, (str)])
-        awu_matrix.append(["api_base", api_base, False, (str)])
-        awu_matrix.append(["api_version", api_version, False, (str)])
-        awu_matrix.append(["engine", engine, False, (str)])
-        awu_matrix.append(["model_name", model_name, False, (str)])
-        # Validate argument types
-        _Validators._validate_function_arguments(awu_matrix)
-
-        # Set API type.
-        self.api_type = api_type
-        # Set API Base URL as follow 
-        # "https://****instance.openai.azure.com/".
-        self.api_base = api_base
-        # Set API version like follow "2022-06-10".
-        self.api_version = api_version
-        # Set API key.
-        self.api_key = api_key
-        # Set LLM engine name.
-        self.__engine = engine
-        # Set model name.
-        self.model_name = model_name
-
-        # Update environment and openai variables.
-        self.__set_llm_env()
-
-        # Initialize AzureOpenAI LLM. 
-        self._llm = AzureOpenAI(engine=self.__engine, model_name=self.model_name)
-
-
-    def __set_llm_env(self):
-        """
-        DESCRIPTION:
-            Internal function to set all LLM info in OS environment and 
-            openai variables.
-            
-        PARAMETERS:
-            None
-        
-        RETURNS:
-            None
-
-        RAISES:
-           None
-
-        EXAMPLES:
-            self.__set_llm_env()
-        """
-        # Set API type.
-        openai.api_type = self.api_type
-        # Set API Base URL as follow 
-        openai.api_base = self.api_base 
-        # Set API version.
-        openai.api_version = self.api_version
-        # Set API key.
-        openai.api_key = self.api_key 
-        # Update OS environment variables.
-        _set_openAI(api_type=self.api_type, api_base=self.api_base, 
-                    api_version=self.api_version, api_key=self.api_key)
-        
-        
-    def get_llm(self):
-        """
-        DESCRIPTION:
-            Get LLM inference endpoint.
-            
-        PARAMETERS:
-            None
-        
-        RETURNS:
-            LLM endpoint object.
-
-        RAISES:
-           None
-
-        EXAMPLES:
-            tdml_ai_obj.get_llm()
-        """
-        return self._llm
-    
-
-    def answer(self, query):
-        """
-        DESCRIPTION:
-            Get the answer to the query.
-            
-        PARAMETERS:
-            query:
-                Required Argument.
-                Specifies the question which needs to be answered by LLM.
-                Types: str
-        
-        RETURNS:
-            str
-
-        RAISES:
-            TeradataMlException, ValueError
-
-        EXAMPLES:
-            tdml_ai_obj.run("Tell me a joke")
-        """
-        awu_matrix = []
-        awu_matrix.append(["query", query, False, (str)])
-        # Validate argument types
-        _Validators._validate_function_arguments(awu_matrix)
-        return self._llm(query)
-    
-
-class DBChain(tdmlAI):
-    """Class manages database chain and communication between user, database, and LLM."""
-    # Prompt to retrieve the answer with high accuracy.
-    __DEFAULT_TEMPLATE = """
-        You are a Teradata Vantage DataBase expert. Given an input question, 
-        first create a syntactically correct Teradata query to run, 
-        then look at the results of the query and return the answer to 
-        the input question. Unless the user specifies in the question a specific 
-        number of examples to obtain, query for at most 2 results using the LIMIT 
-        clause as per Teradata SQL query. You can order the results to return the most 
-        informative data in the database. Never query for all columns from a table. 
-        You must query only the columns that are needed to answer the question. 
-        Wrap each column name in double quotes (") to denote them as delimited identifiers.
-        
-        Pay attention to use only the column names you can see in the tables below. 
-        Be careful to not query for columns that do not exist. Also, pay attention 
-        to which column is in which table.
-        Pay attention to select correct table names and column names for teradatasql query generation.
-
-        Given an input question, first create a syntactically correct teradatasql query 
-        to run, then look at the results of the query and return the answer.
-        Use the following format:
-
-        Question: "Question here"
-        SQLQuery: "SQL Query to run"
-        SQLResult: "Result of the SQLQuery"
-        Answer: "Final answer here"
-
-        Only use the following tables:
-
-        {table_info}
-
-        If someone asks to list out tables present in 'DBC' DataBase user, 
-        then following is an example teradatasql query for 'DBC' user:
-        SQL Query:      SELECT  TableName,
-                        FROM    DBC.TablesV
-                        WHERE   TableKind = 'T'
-                        and     DatabaseName = 'DBC'
-                        ORDER BY  TableName; 
-
-
-        Question: {input}"""
-
-    def __init__(self, llm, include_tables=None, verbose=False, ignore_tdml_prompt=False):        
-        """
-        DESCRIPTION:
-            Constructor of DBChain that sets up interaction channel between client (user), 
-            Vantage database, and LLM endpoint.
-            Notes:
-                * LLM input tokens size must be greater than 1k to perform basic operations
-                  and increase LLM token size based on operation complexity.
-                * Establish connection with Vantage before initializing DBChain.
-            
-        PARAMETERS:
-            llm:
-                Required Argument.
-                Specifies the LLM endpoint object for inference.
-                Types: tdmlAI
-
-            include_tables:
-                Optional Argument.
-                Specifies the table names to be included for interaction.
-                When a list of tables is provided, DBChain gives the 
-                highest priority to those tables for data exploration.
-                Otherwise, entire database tables present in Vantage is taken 
-                into account for data exploration.
-                Notes:
-                    * List of tables has to be provided for more 
-                      accurate results.
-                    * Provide list of tables When LLM input token size is 
-                      less than 4k.
-                    * Views present in database may cause inconsistent result when
-                      include_tables are not provided.
-                Types: list of str
-            
-            verbose:
-                Optional Argument.
-                Specifies whether to display the interaction between LLM and Vantage.
-                Default Value: False
-                Types: bool
-
-            ignore_tdml_prompt:
-                Optional Argument.
-                Specifies wether to ignore tdml engineered prompt design.
-                When "ignore_tdml_prompt" is set to True then tdml engineered 
-                prompt is ignored. Otherwise, prompt is used for accurate results.
-                Notes:
-                    * Custom engineered prompt is used for higher accurate results.
-                    * Set "ignore_tdml_prompt" to True when LLM contains limitation on 
-                      input token size.
-                Default Value: False
-                Types: bool
-
-        RETURNS:
-            TeradataMlException, ValueError, TypeError
-
-        RAISES:
-           None
-
-        EXAMPLES:
-            # Import the modules.
-            from teradataml.gen_ai.convAI import tdmlAI
-            # Create LLM endpoint.
-            tdml_ai_obj = tdmlAI(api_type = "azure",
-                     api_base = "https://****.openai.azure.com/",
-                     api_version = "2000-11-35",
-                     api_key = <provide your llm API key>,
-                     engine=<provide your llm engine name>,
-                     model_name="gpt-3.5-turbo")
-
-            # Create DBChain object.
-            dbchain_obj = DBChain(llm=tdml_ai_obj,verbose=False)
-        """
-
-        # Argument validations
-        awu_matrix = []
-        awu_matrix.append(["llm", llm, False, (tdmlAI)])
-        awu_matrix.append(["include_tables", include_tables, True, (list)])
-        awu_matrix.append(["verbose", verbose, True, (bool)])
-        awu_matrix.append(["ignore_tdml_prompt", ignore_tdml_prompt, True, (bool)])
-        # Validate argument types
-        _Validators._validate_function_arguments(awu_matrix)
-        # Set LLM inference endpoint used for sql query generation.
-        self._llm = llm.get_llm()
-        # Set table names needs to be explored.
-        self._table_names = include_tables
-        # Set verbose to display intermediate steps.
-        self._verbose = verbose
-        # Database custom prompt to improve the SQL generation accuracy.
-        self.__PROMPT = PromptTemplate(
-            input_variables=["input", "table_info"], template=self.__DEFAULT_TEMPLATE ) \
-                if not ignore_tdml_prompt else None
-        # Maintains interaction records.
-        self.__records = []
-
-        # Get all table names except views when table name is 'None'.
-        if self._table_names is None:
-            self.__pull_db_details()
-        # set tdsqlalchemy engine.
-        self.__engine = get_context()
-
-        # Initialize SQLDatabase chain.
-        self.__db = SQLDatabase(engine=self.__engine, include_tables=self._table_names)
-        # Initialize SQLDatabase Chain.
-        self.__db_chain = SQLDatabaseChain.from_llm(self._llm, self.__db, verbose=verbose, prompt=self.__PROMPT)
-
-
-    def run(self, query):
-        """
-        DESCRIPTION:
-            Function interact with Vantage Database using human natural 
-            language.
-            
-        PARAMETERS:
-            query:
-                Required Argument.
-                Specifies the question which needs to be answered by LLM.
-                Query must be precise and represented in English.
-                Types: str
-        
-        RETURNS:
-            str
-
-        RAISES:
-            TeradataMlException, ValueError
-
-        EXAMPLES:
-            dbchain_obj.run("How many house present in Boston")
-        """
-        awu_matrix = []
-        awu_matrix.append(["query", query, False, (str)])
-        # Validate argument types
-        _Validators._validate_function_arguments(awu_matrix)
-
-        # Trigger the Database chain interaction.
-        result = self.__db_chain.run(query)
-        self.__records.append({query:result})
-        # return result when verbose is set to 'False'.
-        if not self._verbose:
-            return result
-
-    
-    def __pull_db_details(self):
-        """
-        DESCRIPTION:
-            Internal function to retrieve table names.
-
-        PARAMETERS:
-            None
-
-        RETURNS:
-            list
-
-        RAISES:
-            None
-
-        EXAMPLES:
-            self.__pull_db_details()
-        """
-        _connection = get_connection()
-        self._table_names = _connection.dialect.get_table_names(_connection)
-        
-    
-    def get_interaction_records(self):
-        """
-        DESCRIPTION:
-            Get the interaction records.
-
-        PARAMETERS:
-            None
-
-        RETURNS:
-            list
-
-        RAISES:
-            None
-
-        EXAMPLES:
-            dbchain_obj.get_interaction_records()
-        """
-        return self.__records
+# ##################################################################
+#
+# Copyright 2023 Teradata. All rights reserved.
+# TERADATA CONFIDENTIAL AND TRADE SECRET
+#
+# Primary Owner: Kesavaragavan B (kesavaragavan.b@teradata.com)
+# Secondary Owner: Prafulla V Tekawade (prafulla.tekawade@teradata.com)
+#
+# This file implements tdmlAI and DBChain which is used for Vantage 
+# database interaction. The tdmlAI manages LLM inference endpoints and
+# DBChain enables interaction between user, Vantage database.
+# Notes: 
+#   * This code is only for internal use. 
+#   * The code may perform modify, create, or delete operations 
+#     in database based on given query. Hence, limit the permissions 
+#     granted to the credentials.
+# 
+# ##################################################################
+
+# Import required packages.
+import openai
+import os
+from langchain.llms import AzureOpenAI
+from langchain.utilities import SQLDatabase
+from langchain_experimental.sql import SQLDatabaseChain
+from langchain.prompts.prompt import PromptTemplate
+from teradataml import get_context, get_connection
+from teradataml.utils.validators import _Validators
+
+def _set_openAI(api_key, api_type, api_base, api_version):
+    """
+        DESCRIPTION:
+            Internal function to set environment variables for AzureAI.
+            
+        PARAMETERS:
+            api_key:
+                Required Argument.
+                Specifies the AzureAI API key.
+                Types: str
+
+            api_type:
+                Required Argument.
+                Specifies the AzureAI API type.
+                Types: str
+            
+            api_base:
+                Required Argument.
+                Specifies the AzureAI API base url.
+                Types: str
+
+            api_version:
+                Required Argument.
+                Specifies the AzureAI API version.
+                Types: str
+
+        RETURNS:
+            None
+
+        RAISES:
+           None
+
+        EXAMPLES:
+            _set_openAI(api_type = "azure"
+                        api_base = "https://***.openai.azure.com/"
+                        api_version = "2021-12-35"
+                        api_key = "999***")
+        """
+    # Set API type.
+    os.environ["OPENAI_API_TYPE"] = api_type
+    # Set API version like follow "2022-06-10".
+    os.environ["OPENAI_API_VERSION"] = api_version
+    # Set API Base URL as follow 
+    # "https://****instance.openai.azure.com/".
+    os.environ["OPENAI_API_BASE"] = api_base
+    # Set API key.
+    os.environ["OPENAI_API_KEY"] = api_key     
+
+
+class tdmlAI:
+    """tdmlAI provides convenient access to the LLM endpoints for inference."""
+
+    def __init__(self,api_key, 
+                 api_type,
+                 api_base, 
+                 api_version, 
+                 engine,
+                 model_name):
+        """
+        DESCRIPTION:
+            Constructor of tdmlAI that sets up the environment and 
+            initializes the LLM endpoint.
+            
+        PARAMETERS:
+            api_key:
+                Required Argument.
+                Specifies the LLM API key.
+                Types: str
+
+            api_type:
+                Required Argument.
+                Specifies the LLM API type.
+                Types: str
+            
+            api_base:
+                Required Argument.
+                Specifies the LLM API base url.
+                Types: str
+
+            api_version:
+                Required Argument.
+                Specifies the LLM API version.
+                Types: str                       
+                  
+            engine:
+                Required Argument.
+                Specifies the deployment name of the LLM.
+                Types: str           
+
+            model_name:
+                Required Argument.
+                Specifies the LLM model name.
+                Types: str
+
+        RETURNS:
+            None
+
+        RAISES:
+            TeradataMlException, ValueError, TypeError
+
+        EXAMPLES:
+            # Import the modules.
+            from teradataml.gen_ai.convAI import tdmlAI
+            # Create LLM endpoint.
+            tdml_ai_obj = tdmlAI(api_type = "azure",
+                     api_base = "https://****.openai.azure.com/",
+                     api_version = "2000-11-35",
+                     api_key = <provide your llm API key>,
+                     engine = <provide your llm engine name>,
+                     model_name = "gpt-3.5-turbo")
+        """
+
+        # Argument validations
+        awu_matrix = []
+        awu_matrix.append(["api_key", api_key, False, (str)])
+        awu_matrix.append(["api_type", api_type, False, (str)])
+        awu_matrix.append(["api_base", api_base, False, (str)])
+        awu_matrix.append(["api_version", api_version, False, (str)])
+        awu_matrix.append(["engine", engine, False, (str)])
+        awu_matrix.append(["model_name", model_name, False, (str)])
+        # Validate argument types
+        _Validators._validate_function_arguments(awu_matrix)
+
+        # Set API type.
+        self.api_type = api_type
+        # Set API Base URL as follow 
+        # "https://****instance.openai.azure.com/".
+        self.api_base = api_base
+        # Set API version like follow "2022-06-10".
+        self.api_version = api_version
+        # Set API key.
+        self.api_key = api_key
+        # Set LLM engine name.
+        self.__engine = engine
+        # Set model name.
+        self.model_name = model_name
+
+        # Update environment and openai variables.
+        self.__set_llm_env()
+
+        # Initialize AzureOpenAI LLM. 
+        self._llm = AzureOpenAI(engine=self.__engine, model_name=self.model_name)
+
+
+    def __set_llm_env(self):
+        """
+        DESCRIPTION:
+            Internal function to set all LLM info in OS environment and 
+            openai variables.
+            
+        PARAMETERS:
+            None
+        
+        RETURNS:
+            None
+
+        RAISES:
+           None
+
+        EXAMPLES:
+            self.__set_llm_env()
+        """
+        # Set API type.
+        openai.api_type = self.api_type
+        # Set API Base URL as follow 
+        openai.api_base = self.api_base 
+        # Set API version.
+        openai.api_version = self.api_version
+        # Set API key.
+        openai.api_key = self.api_key 
+        # Update OS environment variables.
+        _set_openAI(api_type=self.api_type, api_base=self.api_base, 
+                    api_version=self.api_version, api_key=self.api_key)
+        
+        
+    def get_llm(self):
+        """
+        DESCRIPTION:
+            Get LLM inference endpoint.
+            
+        PARAMETERS:
+            None
+        
+        RETURNS:
+            LLM endpoint object.
+
+        RAISES:
+           None
+
+        EXAMPLES:
+            tdml_ai_obj.get_llm()
+        """
+        return self._llm
+    
+
+    def answer(self, query):
+        """
+        DESCRIPTION:
+            Get the answer to the query.
+            
+        PARAMETERS:
+            query:
+                Required Argument.
+                Specifies the question which needs to be answered by LLM.
+                Types: str
+        
+        RETURNS:
+            str
+
+        RAISES:
+            TeradataMlException, ValueError
+
+        EXAMPLES:
+            tdml_ai_obj.run("Tell me a joke")
+        """
+        awu_matrix = []
+        awu_matrix.append(["query", query, False, (str)])
+        # Validate argument types
+        _Validators._validate_function_arguments(awu_matrix)
+        return self._llm(query)
+    
+
+class DBChain(tdmlAI):
+    """Class manages database chain and communication between user, database, and LLM."""
+    # Prompt to retrieve the answer with high accuracy.
+    __DEFAULT_TEMPLATE = """
+        You are a Teradata Vantage DataBase expert. Given an input question, 
+        first create a syntactically correct Teradata query to run, 
+        then look at the results of the query and return the answer to 
+        the input question. Unless the user specifies in the question a specific 
+        number of examples to obtain, query for at most 2 results using the LIMIT 
+        clause as per Teradata SQL query. You can order the results to return the most 
+        informative data in the database. Never query for all columns from a table. 
+        You must query only the columns that are needed to answer the question. 
+        Wrap each column name in double quotes (") to denote them as delimited identifiers.
+        
+        Pay attention to use only the column names you can see in the tables below. 
+        Be careful to not query for columns that do not exist. Also, pay attention 
+        to which column is in which table.
+        Pay attention to select correct table names and column names for teradatasql query generation.
+
+        Given an input question, first create a syntactically correct teradatasql query 
+        to run, then look at the results of the query and return the answer.
+        Use the following format:
+
+        Question: "Question here"
+        SQLQuery: "SQL Query to run"
+        SQLResult: "Result of the SQLQuery"
+        Answer: "Final answer here"
+
+        Only use the following tables:
+
+        {table_info}
+
+        If someone asks to list out tables present in 'DBC' DataBase user, 
+        then following is an example teradatasql query for 'DBC' user:
+        SQL Query:      SELECT  TableName,
+                        FROM    DBC.TablesV
+                        WHERE   TableKind = 'T'
+                        and     DatabaseName = 'DBC'
+                        ORDER BY  TableName; 
+
+
+        Question: {input}"""
+
+    def __init__(self, llm, include_tables=None, verbose=False, ignore_tdml_prompt=False):        
+        """
+        DESCRIPTION:
+            Constructor of DBChain that sets up interaction channel between client (user), 
+            Vantage database, and LLM endpoint.
+            Notes:
+                * LLM input tokens size must be greater than 1k to perform basic operations
+                  and increase LLM token size based on operation complexity.
+                * Establish connection with Vantage before initializing DBChain.
+            
+        PARAMETERS:
+            llm:
+                Required Argument.
+                Specifies the LLM endpoint object for inference.
+                Types: tdmlAI
+
+            include_tables:
+                Optional Argument.
+                Specifies the table names to be included for interaction.
+                When a list of tables is provided, DBChain gives the 
+                highest priority to those tables for data exploration.
+                Otherwise, entire database tables present in Vantage is taken 
+                into account for data exploration.
+                Notes:
+                    * List of tables has to be provided for more 
+                      accurate results.
+                    * Provide list of tables When LLM input token size is 
+                      less than 4k.
+                    * Views present in database may cause inconsistent result when
+                      include_tables are not provided.
+                Types: list of str
+            
+            verbose:
+                Optional Argument.
+                Specifies whether to display the interaction between LLM and Vantage.
+                Default Value: False
+                Types: bool
+
+            ignore_tdml_prompt:
+                Optional Argument.
+                Specifies wether to ignore tdml engineered prompt design.
+                When "ignore_tdml_prompt" is set to True then tdml engineered 
+                prompt is ignored. Otherwise, prompt is used for accurate results.
+                Notes:
+                    * Custom engineered prompt is used for higher accurate results.
+                    * Set "ignore_tdml_prompt" to True when LLM contains limitation on 
+                      input token size.
+                Default Value: False
+                Types: bool
+
+        RETURNS:
+            TeradataMlException, ValueError, TypeError
+
+        RAISES:
+           None
+
+        EXAMPLES:
+            # Import the modules.
+            from teradataml.gen_ai.convAI import tdmlAI
+            # Create LLM endpoint.
+            tdml_ai_obj = tdmlAI(api_type = "azure",
+                     api_base = "https://****.openai.azure.com/",
+                     api_version = "2000-11-35",
+                     api_key = <provide your llm API key>,
+                     engine=<provide your llm engine name>,
+                     model_name="gpt-3.5-turbo")
+
+            # Create DBChain object.
+            dbchain_obj = DBChain(llm=tdml_ai_obj,verbose=False)
+        """
+
+        # Argument validations
+        awu_matrix = []
+        awu_matrix.append(["llm", llm, False, (tdmlAI)])
+        awu_matrix.append(["include_tables", include_tables, True, (list)])
+        awu_matrix.append(["verbose", verbose, True, (bool)])
+        awu_matrix.append(["ignore_tdml_prompt", ignore_tdml_prompt, True, (bool)])
+        # Validate argument types
+        _Validators._validate_function_arguments(awu_matrix)
+        # Set LLM inference endpoint used for sql query generation.
+        self._llm = llm.get_llm()
+        # Set table names needs to be explored.
+        self._table_names = include_tables
+        # Set verbose to display intermediate steps.
+        self._verbose = verbose
+        # Database custom prompt to improve the SQL generation accuracy.
+        self.__PROMPT = PromptTemplate(
+            input_variables=["input", "table_info"], template=self.__DEFAULT_TEMPLATE ) \
+                if not ignore_tdml_prompt else None
+        # Maintains interaction records.
+        self.__records = []
+
+        # Get all table names except views when table name is 'None'.
+        if self._table_names is None:
+            self.__pull_db_details()
+        # set tdsqlalchemy engine.
+        self.__engine = get_context()
+
+        # Initialize SQLDatabase chain.
+        self.__db = SQLDatabase(engine=self.__engine, include_tables=self._table_names)
+        # Initialize SQLDatabase Chain.
+        self.__db_chain = SQLDatabaseChain.from_llm(self._llm, self.__db, verbose=verbose, prompt=self.__PROMPT)
+
+
+    def run(self, query):
+        """
+        DESCRIPTION:
+            Function interact with Vantage Database using human natural 
+            language.
+            
+        PARAMETERS:
+            query:
+                Required Argument.
+                Specifies the question which needs to be answered by LLM.
+                Query must be precise and represented in English.
+                Types: str
+        
+        RETURNS:
+            str
+
+        RAISES:
+            TeradataMlException, ValueError
+
+        EXAMPLES:
+            dbchain_obj.run("How many house present in Boston")
+        """
+        awu_matrix = []
+        awu_matrix.append(["query", query, False, (str)])
+        # Validate argument types
+        _Validators._validate_function_arguments(awu_matrix)
+
+        # Trigger the Database chain interaction.
+        result = self.__db_chain.run(query)
+        self.__records.append({query:result})
+        # return result when verbose is set to 'False'.
+        if not self._verbose:
+            return result
+
+    
+    def __pull_db_details(self):
+        """
+        DESCRIPTION:
+            Internal function to retrieve table names.
+
+        PARAMETERS:
+            None
+
+        RETURNS:
+            list
+
+        RAISES:
+            None
+
+        EXAMPLES:
+            self.__pull_db_details()
+        """
+        _connection = get_connection()
+        self._table_names = _connection.dialect.get_table_names(_connection)
+        
+    
+    def get_interaction_records(self):
+        """
+        DESCRIPTION:
+            Get the interaction records.
+
+        PARAMETERS:
+            None
+
+        RETURNS:
+            list
+
+        RAISES:
+            None
+
+        EXAMPLES:
+            dbchain_obj.get_interaction_records()
+        """
+        return self.__records