COMPUTING PLATFORM FOR NEURO-SYMBOLIC ARTIFICIAL INTELLIGENCE APPLICATIONS
A distributed generative artificial intelligence (AI) reasoning and action platform that utilizes a cloud-based computing architecture for neuro-symbolic reasoning. The platform comprises systems for distributed computation, curation, marketplace integration, and context management. A distributed computational graph (DCG) orchestrates complex workflows for building and deploying generative AI models, incorporating expert judgment and external data sources. A context computing system aggregates contextual data, while a curation system provides curated responses from trained models. Marketplaces offer data, algorithms, and expert judgment for purchase or integration. The platform enables enterprises to construct user-defined workflows and incorporate trained models into their business processes, leveraging enterprise-specific knowledge. The platform facilitates flexible and scalable integration of machine learning models into software applications, supported by a dynamic and adaptive DCG architecture.
1 . A computing system comprising:
at least a memory and one or more hardware processors configured to:
obtain a plurality of input data, the input data comprising enterprise knowledge and expert knowledge;
process the input data using an embedding engine to generate a vectorized dataset;
apply a machine learning model to the vectorized dataset, the machine learning model having been trained to identify relationships among the input data, to generate a parameter set;
map the parameter set to a plurality of symbolic rules;
generate a structured execution graph representing a course of action by applying a symbolic reasoning engine to the symbolic rules;
determine, for at least a subset of nodes in the structured execution graph, one or more physical computing locations for execution; and
initiate execution of the structured execution graph at the one or more determined physical computing locations.
2 . The computing system of claim 1 , further comprising a model interaction interface configured to:
automatically generate a natural language query based on at least a portion of the symbolic rules;
transmit the query to a large language model; and
receive a response for modifying, supplementing, or validating one or more nodes of the structured execution graph.
3 . The computing system of claim 1 , wherein the symbolic reasoning engine incorporates retrieval-augmented generation (RAG) information into the execution of one or more of the symbolic rules.
4 . The computing system of claim 1 , wherein the symbolic reasoning engine generates the structured execution graph based in part on user-specific information derived from enterprise knowledge or prior interactions.
5 . The computing system of claim 1 , wherein determining the physical computing locations comprises evaluating device-specific locality information associated with a user device.
6 . The computing system of claim 1 , wherein the expert knowledge comprises scored datasets or scored model output.
7 . The computing system of claim 6 , wherein the expert knowledge is obtained from an expert knowledge marketplace.
8 . A computer-implemented method comprising the steps of:
obtaining a plurality of input data, the input data comprising enterprise knowledge and expert knowledge;
processing the obtained plurality of input data using an embedding model to create a vectorized dataset;
obtaining a plurality of input data, the input data comprising enterprise knowledge and expert knowledge;
processing the input data using an embedding engine to generate a vectorized dataset;
applying a machine learning model to the vectorized dataset, the machine learning model having been trained to identify relationships among the input data, to generate a parameter set;
mapping the parameter set to a plurality of symbolic rules;
generate a structured execution graph representing a course of action by applying a symbolic reasoning engine to the plurality of symbolic rules;
determining, for at least a subset of nodes of the structured execution graph, one or more physical computing locations for execution based on locality constraints, regulatory limitations, or available resources; and
initiating execution of the structured execution graph at the one or more determined physical computing locations.
9 . The computer-implemented method of claim 8 , further comprising:
generating a natural language query based on at least a portion of the symbolic rules;
transmitting the query to a large language model; and
receiving a response for modifying, supplementing, or validating one or more nodes of the structured execution graph.
10 . The computer-implemented method of claim 98 , wherein the symbolic reasoning engine incorporates retrieval-augmented generation (RAG) information to inform one or more symbolic rules.
11 . The computer-implemented method of claim 8 , wherein the symbolic reasoning engine accounts for information associated with an action a user of the user device is performing during interaction with the platform.
12 . The computer-implemented method of claim 8 wherein determining the physical computing locations comprises evaluating device-specific locality information associated with a user device.
13 . The computer-implemented method of claim 8 , wherein the expert knowledge comprises scored datasets or scored model output.
14 . The computer-implemented method of claim 13 , wherein the expert knowledge is obtained from an expert knowledge marketplace.
15 . A system comprising one or more computers each with a memory and at least one processor and executable instructions that, when executed on one or more of the computers, cause the system to:
obtain a plurality of input data, the input data comprising enterprise knowledge and expert knowledge;
process the obtained plurality of input data using an embedding model to create a vectorized dataset;
apply a machine learning model to the vectorized dataset, the machine learning model having been trained to identify relationships among the input data, to generate a parameter set;
map the parameter set to a plurality of symbolic rules;
generate a structured execution graph representing a course of action by applying a symbolic reasoning engine to the plurality of symbolic rules;
determine, for at least a subset of nodes in the structured execution graph, one or more physical computing locations for execution based on locality constraints, regulatory limitations, or available resources; and
initiate execution of the structured execution graph at the one or more determined physical computing locations.
16 . The system of claim 15 , further comprising a model interaction interface configured to:
generate a natural language query based on at least a portion of the symbolic rules;
transmit the query to a large language model; and
receive a response for modifying, supplementing, or validating one or more nodes of the structured execution graph.
17 . The system of claim 15 , wherein the symbolic reasoning engine incorporates retrieval-augmented generation (RAG) information to inform one or more symbolic rules.
18 . The system of claim 15 , wherein the symbolic reasoning engine accounts for information associated with an action a user of the user device is performing during interaction with the platform.
19 . The system of claim 15 , wherein determining the physical computing locations comprises evaluating device-specific locality information associated with a user device.
20 . The system of claim 15 , wherein the expert knowledge comprises scored datasets or scored model output.
21 . The system of claim 20 , wherein the expert knowledge is obtained from an expert knowledge marketplace.
22 . Non-transitory, computer-readable storage media having computer executable instructions embodied thereon that, when executed by one or more processors of a computing system, cause the computing system to:
obtain a plurality of input data, the input data comprising enterprise knowledge and expert knowledge;
process the obtained plurality of input data using an embedding model to create a vectorized dataset;
apply a machine learning model to the vectorized dataset, the machine learning model having been trained to identify relationships among the input data, to generate a parameter set;
map the parameter set to a plurality of symbolic rules;
generate a structured execution graph representing a course of action by applying a symbolic reasoning engine to the symbolic rules;
determine, for at least a subset of nodes in the structured execution graph, one or more physical computing locations for execution based on locality constraints, regulatory limitations, or available resources; and
initiate execution of the structured execution graph at the one or more determined physical computing locations.
23 . The system of claim 22 , further comprising a model interaction interface configured to:
generate a natural language query based on at least a portion of the symbolic rules;
transmit the query to a large language model; and
receive a response for modifying, supplementing, or validating one or more nodes of the structured execution graph.
24 . The system of claim 22 , wherein the symbolic reasoning engine incorporates retrieval augmented generation (RAG) information to inform one or more symbolic rules.
25 . The system of claim 22 , wherein the symbolic reasoning engine accounts for information associated with an action a user of the user device is performing during interaction with the platform.
26 . The system of claim 22 , wherein determining the physical computing locations comprises evaluating device-specific locality information associated with a user device.
27 . The system of claim 22 , wherein the expert knowledge comprises scored datasets or scored model output.
28 . The system of claim 27 , wherein the expert knowledge is obtained from an expert knowledge marketplace.