A computer-implemented method executed by a controller of an integrated entity analysis system for structuring, indexing, retrieving, and analyzing entity data for automating asset trading, the method comprising:
- for each entity of a plurality of entities:
- configuring, by the controller, an AI model to generate one or more natural-language search strings based on information missing from an entity profile of the entity;
- retrieving, by the controller, unstructured external data related to the entity from unstructured data sources using the natural-language search strings;
- applying, by the controller, the unstructured external data to the AI model to generate structured property/value pairs describing the characteristics of the entity;
- classifying, by the controller using the AI model, the entity based on the structured property/value pairs;
- selecting, by the controller, an entity-profile template associated with the classification;
- mapping, by the controller, the structured property/value pairs into corresponding fields of the entity-profile template and, when a property lacks a corresponding field, dynamically generating a new field in the entity profile and populating the new field with the corresponding value;
- storing, by the controller, the updated entity profile in an entity database;
- and
- converting, by the controller, the updated entity profile into a vector embedding and maintaining the vector embedding in a structured retrieval dataset comprising vector embeddings for a plurality of entity profiles;
- receiving from a user, by the controller, investor-specified risk parameters and portfolio criteria;
- converting, by the controller, a scoring-attribute definition or a user query into a vector embedding, and performing a similarity search across the structured retrieval dataset to identify portions of the entity profiles relevant to the scoring-attribute definition or the user query;
- supplying, by the controller, the identified relevant portions of the entity profiles as context to the AI model, and generating, by the AI model, scoring data for the plurality of entity profiles based on the context;
- based on scoring preferences of the user, selecting, by the controller, from the plurality of entity profiles using the scoring data, a first set of entity profiles corresponding to candidate assets that satisfy the investor-specified risk parameters and the portfolio criteria;
- constructing, by the controller, a portfolio based on the investor-specified risk parameters, including the risk tolerance of the investor, and the portfolio criteria, the constructing including, for each candidate asset associated with an entity profile in the first set of entity profiles, determining based on the investor-specified risk parameters and the portfolio criteria, by the controller, an asset allocation for the portfolio;
- and
- transmitting, by the controller, instructions to one or more trade fulfillment backends to perform one or more trades in accordance with the determined asset allocations,
- wherein generating the structured property/value pairs, dynamically generating fields in the entity profile, converting the entity profiles into vector embeddings, maintaining the structured retrieval dataset, and performing similarity searches to supply context to the AI model reduces processing complexity, reduces computational overhead, improves scalability, and reduces energy consumption in the integrated entity analysis system.