US 11553055

US Patent 11553055, titled "Automated communication-based intelligence engine," was filed by Numberai Inc on January 7, 2021, and issued on January 10, 2023. The inventors listed are Daniel Hodges, Tasos Roumeliotis, Joel Grossman, Andrew Ruff, Steven Ginn, and Andrew Smith.

Abstract:
The patent describes a system that automatically creates and maintains one or more digital "entity models" based on communication data received over a network. This process utilizes machine learning, a user interface, a sentiment analyzer, a communication monitoring agent, and an automated bot. The bots are built from templates. The system continuously analyzes new communication data to improve the entity model, keep the bot updated, and update other services the entity uses. It also helps businesses categorize other entities to identify competitors and provides analytics on how users interact with the entity model. The system can automatically maintain listings like Frequently Asked Questions (FAQs) and is designed for a networked, potentially distributed, environment.

Independent Claims Overview:

• Claim 1 (System for Entity Model Creation and Updating): This claim describes a computer system with one or more processors programmed to automatically build and maintain an "entity model" for a "primary entity" (e.g., a business). It functions by monitoring the primary entity's communications, using this data to create the model, and then continually monitoring new communications. When new relevant communications are detected, the system applies the existing entity model to generate updates, which are then transmitted to a specific platform via a network (e.g., updating a website's FAQs).
• Claim 11 (Method for Entity Model Creation and Updating): This claim outlines a computer-implemented method that mirrors the functionality of Claim 1. It details the steps performed by processors: monitoring a primary entity's communications, building an entity model, subsequently monitoring particular communications, applying the model to these communications to generate updates for an entity datastore, and finally applying these updates to a platform over a network.
• Claim 12 (Method for Routing Online Conversations): This method describes how a system, through its processors, intelligently routes online conversations to optimize for "conversions" (e.g., sales). It involves:
  1. Collecting numerous electronic communication sessions from various "secondary entities" (e.g., customers) and identifying which ones resulted in a conversion.
  2. Training a model to predict the likelihood of conversion based on these historical conversations.
  3. Receiving new communications from a secondary entity during a session.
  4. Using an artificial response system (like a chatbot) to initially respond.
  5. Analyzing the new communications to determine their conversion likelihood using the trained model.
  6. Deciding whether to hand off the communication session from the artificial response system (e.g., to a human agent) based on the determined conversion likelihood.
• Claim 18 (System for Routing Online Conversations): This claim describes a computer system specifically designed to implement the online conversation routing method detailed in Claim 12. It includes processors configured to receive communications, train a conversion likelihood model, receive and respond to new communications using an artificial response system, analyze conversion likelihood, and hand off sessions based on that likelihood.
• Claim 19 (Method for Entity Categorization Scheme): This claim describes a computer-implemented method for categorizing "secondary entities." The method involves:
  1. Monitoring a multitude of communications involving various secondary entities.
  2. Analyzing these communications to determine characteristics such as the number, timing, length, distribution channel, or content of the communications.
  3. Clustering (grouping) these secondary entities together into at least one cluster based on the analysis.

USPTO and CAFC Docket Search:

A search for US11553055B2 on the USPTO database indicates that the patent is Active and is projected to expire on July 16, 2038.

A search for "CAFC 2026 dockets 11553055" did not return any specific dockets for patent 11553055 in the year 2026 at the Court of Appeals for the Federal Circuit.




https://patents.google.com/patent/US11553055B2/en
https://cafc.uscourts.gov/opinions-orders/