Systems, methods, and graphical user interfaces for secure execution of analytical tasks using natural language

A computer-implemented method comprising:

• initializing a language model according to a set of predefined prompts;
• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by the language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server based at least in part on the initializing, wherein translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input, wherein:
• the one or more input arguments of the analytical function call includes a first input argument,
• the analytical function at the analytics computing server includes a second input argument and a third input argument, different from the second input argument, and
• a value of the second input argument and a value of the third input argument is detected from the first input argument;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output.

A computer-implemented system comprising:

• one or more processors;
• a memory;
• a non-transitory computer-readable medium operably coupled to the one or more processors, the non-transitory computer-readable medium having computer-readable instructions stored thereon that, when executed by the one or more processors, cause a computing device to perform operations comprising:
• initializing a language model according to a set of predefined prompts;
• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by the language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server based at least in part on the initializing, wherein translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input, wherein:
• the one or more input arguments of the analytical function call includes a first input argument,
• the analytical function at the analytics computing server includes a second input argument and a third input argument, different from the second input argument, and
• a value of the second input argument and a value of the third input argument is detected from the first input argument;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output.

A computer-program product comprising a non-transitory machine-readable storage medium storing computer instructions that, when executed by one or more processors, perform operations comprising:

• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by a language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server, wherein:
• each analytical function of the plurality of distinct analytical functions maps to a respective set of predefined instructions accessible by the analytics computing server, and wherein calling the analytical function comprises providing an identifier of the analytical function call to the analytics computing server and executing, at the analytics computing server, the respective set of predefined instructions mapped to the analytical function based at least in part on the identifier of the analytical function call, and
• translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input, wherein:
• the one or more input arguments of the analytical function call include a first input argument and a second input argument, and
• the analytical function at the analytics computing server includes a third input argument and a fourth input argument;
• calling the analytical function at the analytics computing server using the analytical function call, wherein the analytics computing server executes the analytical function based on a value of the third input argument and a value of the fourth input argument, wherein executing the analytical function at the analytics computing server includes:
• using the value detected for the first input argument as the value for the third input argument, and
• using the value detected for the second input argument as the value for the fourth input argument;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output.

A computer-program product comprising a non-transitory machine-readable storage medium storing computer instructions that, when executed by one or more processors, perform operations comprising:

• initializing a language model according to a set of predefined prompts;
• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by the language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server based at least in part on the initializing, wherein translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input, wherein:
• the one or more input arguments of the analytical function call includes a first input argument,
• the analytical function at the analytics computing server includes a second input argument and a third input argument, different from the second input argument, and
• a value of the second input argument and a value of the third input argument is detected from the first input argument;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output.

A computer-program product comprising a non-transitory machine-readable storage medium storing computer instructions that, when executed by one or more processors, perform operations comprising:

• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by a language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server, wherein translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output,
• wherein:
• the one or more input arguments of the analytical function call at least include a required input argument and an optional input argument,
• detecting the value for the one or more input arguments of the analytical function call includes:
• determining that the language model did not detect the value for the required input argument, and
• assigning a default value to the required input argument in response to the determining that the language model did not detect the value for the required input argument,
• the computer instructions, when executed by the one or more processors, perform operations further comprising:
• determining that the value detected for a first input argument of the one or more input arguments is of a first data type,
• determining that the first input argument of the analytical function call corresponds to a second input argument of the analytical function that is of a second data type, different from the first data type,
• converting the value detected for the first input argument from the first data type to the second data type, and
• after converting the value to the second data type, using the value for the second input argument of the analytical function,
• the computer instructions, when executed by the one or more processors, perform operations further comprising:
• translating, by the language model, the technical analytical output to a natural language explanation of the technical analytical output, or
• the natural language input is received via a graphical user interface (GUI), a command-line interface (CLI), or an application programming interface (API).

A computer-implemented method comprising:

• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by a language model that is configured to translate the natural language input to a plurality of distinct analytical function calls that are each associated with a function prompt name and a function prompt description, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server, wherein translating the natural language input to the analytical function call includes:
• detecting, via the language model, a value for one or more input arguments of the analytical function call from the natural language input and that a vector value of the function prompt name and a vector value of the function prompt description of the analytical function call of the plurality of distinct analytical function calls has a smallest vector distance to a vector value of the natural language input, and
• translating the natural language input to the analytical function call based on detecting that the vector value of the function prompt name and the vector value of the function prompt description has the smallest vector distance to the vector value of the natural language input;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output.

A computer-implemented system comprising:

• one or more processors;
• a memory;
• a non-transitory computer-readable medium operably coupled to the one or more processors, the non-transitory computer-readable medium having computer-readable instructions stored thereon that, when executed by the one or more processors, cause a computing device to perform operations comprising:
• receiving a natural language input comprising a natural language request for executing an analytical task;
• processing the natural language input by a language model, wherein the processing includes:
• translating the natural language input to an analytical function call for calling an analytical function of a plurality of distinct analytical functions of an analytics computing server, wherein translating the natural language input to the analytical function call includes detecting a value for one or more input arguments of the analytical function call from the natural language input;
• calling the analytical function at the analytics computing server using the analytical function call;
• receiving a technical analytical output in response to calling the analytical function;
• and
• outputting a response to the natural language input that includes the technical analytical output,
• wherein:
• the one or more input arguments of the analytical function call at least include a required input argument and an optional input argument,
• detecting the value for the one or more input arguments of the analytical function call includes:
• determining that the language model did not detect the value for the required input argument, and
• assigning a default value to the required input argument in response to the determining that the language model did not detect the value for the required input argument,
• the computer-readable instructions stored thereon, when executed by the one or more processors, cause the computing device to perform operations further comprising:
• determining that the value detected for a first input argument of the one or more input arguments is of a first data type,
• determining that the first input argument of the analytical function call corresponds to a second input argument of the analytical function that is of a second data type, different from the first data type,
• converting the value detected for the first input argument from the first data type to the second data type, and
• after converting the value to the second data type, using the value for the second input argument of the analytical function,
• the computer-readable instructions stored thereon, when executed by the one or more processors, cause the computing device to perform operations further comprising:
• translating, by the language model, the technical analytical output to a natural language explanation of the technical analytical output, or
• the natural language input is received via a graphical user interface (GUI), a command-line interface (CLI), or an application programming interface (API).