Palantir Technologies

A system or computerized method may receive a request from a front-end application for visualizing a plurality of data objects on a geospatial map, each of the data objects associated with a corresponding data object type. A system or method may access an ontology that includes associations between data object types and rendering specifications, the rendering specifications being in a standard format for a plurality of front-end applications. A system or method may be based on the rendering specifications associated with the plurality of data objects, generating a plurality of renderables corresponding to each of the plurality of front-end applications. A system or method may provide corresponding renderables to the requesting front-end application that are usable to visualize the plurality of data objects on the geospatial map.

Computer-implemented methods and systems including an agent service and a long-term memory store service that communicates with a long-term memory store comprising a knowledge graph, and that services long-term memory storage and retrieval requests from the agent service. Long-term memory retrieval requests include a query and memory types, and are serviced through determining a resulting set of nodes from the knowledge graph and aggregating associated memory content. Long-term memory storage requests include memory content and memory types, and are serviced through determining related nodes and relationships, and adding a new node with the memory content and new edges to the related nodes based on the determined relationships.

In some examples, systems and methods for using map data are provided. An example method for collaborative map use is provided. In some examples, the method includes receiving a first request of a map of a specific geographic area from a first user device; receiving a first language code that corresponds to a first language preferred by a first user associated with the first user device; selecting a first map tile variant from a plurality of map tile variants based on the first language code; and transmitting an indication of the first map tile variant to the first user device to display a first map of the specific geographic area using the first map tile variant, the first displayed map including one or more first text strings in the first language.

A method may provide a set of data objects associated with an ontology. A method may receive a user query associated with the set of data objects, provide ontology documentation indicative of a method for accessing data objects, wherein the ontology documentation is specific to the ontology, generate a prompt for providing to a large language model ("LLM")-based agent, wherein the prompt includes at least: the ontology documentation and the user query, and provide the prompt to the LLM-based agent to cause the LLM-based agent to at least: generate a search query based on the ontology documentation and the user query, cause execution of the search query to access one or more data objects; and generate a response to the prompt based at least in part on the one or more data objects. A method may receive and causing display of the response in an interactive graphical user interface.

Systems and methods for code rebasing/conflict resolution are provided. For example, method for rebasing a software project is comprises: determining a set of changes between a branch version and a main version of the software project; identifying, from the set of changes, a conflict between the branch version and the main version; generating a user interface comprising at least one of: a first graphical representation of the conflict as implemented in the main version; or a second graphical representation of the conflict as implemented in the branch version; receiving, via the user interface, user input indicating a resolution for the presented conflict; and updating the branch version of the software project according to the indicated resolution, thereby resolving the conflict to facilitate merging of the branch of the software project onto the main version of the software project; wherein the method is performed using one or more processors.

Disclosed herein are computing environments and processes for protocol-agnostic inbound data ingestion to securely and intuitively support a wide and varied range of push-based data integrations. An event detector on a perimeter network can receive inbound data from an external system. A data ingestion engine of a permissioned system can access a repository of the permissioned system, can validate the inbound data based on one or more attributes of data in the repository, and can provide a validation indication to the event detector. The event detector can provide the inbound data to the permissioned system responsive to the validation indication from the data ingestion engine.

Systems and methods for context-aware attachments and large language model (LLM) analyses. An example method includes presenting an assistant application and a development application in a browser, the applications being in communication with a back-end system. Communications are established between the applications using a communication channel. A user prompt and indication of context to be included in an LLM prompt are obtained, with the user prompt and indication of context being routed to the back-end system and the indication of context being specified based on attachment of a repository, a code file being edited in the development application, or a portion of the code file selected in the development application. A response from an LLM is triggered and the response is presented in the assistant application.

A computer-implemented method may include: generating an interactive graphical visualization including a first node representing an automation object, a second node representing an action object, and a third node representing a data object type or a data object; determining one or more conditions of the automation object which, when satisfied, trigger an automation of the automation object; determining that the one or more conditions reference the data object or a property; determining that execution of an action of the action object causes the one or more conditions to be satisfied; updating the interactive graphical visualization to include a first edge connecting the second node and the first node to represent that the execution of the action of the action object can cause the one or more conditions to be satisfied to trigger the automation.

Computer-implemented systems and methods are disclosed, including systems and methods utilizing language models for extracting and resolving entities mentioned in unstructured text data. A computer-implemented method may include at least: receiving first text data from a data source; generating and transmitting a first prompt for a large language model ("LLM"), the first prompt comprising the first text data; receiving a first output from the LLM in response to the first prompt; generating, based at least on the first output, and transmitting a second prompt to the LLM; receiving a second output from the LLM in response to the second prompt, the second output comprising first one or more entities extracted by the LLM from the first part of the first semantically related text portion and first metadata or context associated with the first one or more entities.

Computer-implemented systems and methods are disclosed for transitioning interactive workflow graphical user interface generation to local software development environments. A computer-implemented method may include receiving, via interactive user interface and from a user, definitions of one or more data object types including respective one or more properties. A computer-implemented method may execute a build process on the definitions to generate: an ontology definition file; and an ontology software development kit ("SDK"). A computer-implemented method may automatically communicate the ontology definition file to a backend server that may be usable by the backend server to generate an ontology implementation that may be usable in execution of software applications that operate on data objects. A computer-implemented method may include automatically implementing the ontology SDK in the software development environment, wherein the ontology SDK is configured to provide functionality to the software development environment for developing software applications that operate on data objects.