Description:FIELD OF THE INVENTION
[0001] The present invention relates to enterprise data access and management. The invention focuses on real-time data democratization across distributed environments. The invention specifically involves a transient data marketplace for dynamic querying and decentralized data publishing.
BACKGROUND FOR THE INVENTION:
[0002] Cloud Modern enterprise systems operate across a variety of hardware configurations, including cloud servers, edge devices, and on-premises infrastructure. These systems host structured databases, unstructured repositories, and AI-generated insights, each requiring distinct protocols and interfaces for access. Existing solutions ingest data from data sources while doing data operations. This results in fragmented architecture, increased latency, and high maintenance overhead due to the need for multiple integration layers and specialized hardware configurations.
[0003] Data marketplaces traditionally rely on centralized architectures that struggle to scale across distributed hardware environments. These systems often fail to support concurrent access to diverse data products, leading to performance bottlenecks. Moreover, metadata management and lineage tracking mechanisms are not optimized for real-time updates, especially when executed across multiple processors and memory units. The lack of dynamic visualization and adaptive query processing further limits the responsiveness and usability of these platforms, particularly in high-volume enterprise scenarios.
[0004] Data democratization refers to the technical capability of enabling secure, scalable, and real-time access to diverse data sources across distributed computing environments without centralized ingestion or replication. The invention facilitates this by implementing a unified virtual interface layer that establishes connections to structured databases, unstructured repositories, and external APIs, allowing dynamic data retrieval directly from source systems. Through the data marketplace, users and systems can interact with adaptive data packages generated on-the-fly, supporting concurrent access, query customization, and decentralized publishing. This architecture eliminates fragmentation and latency associated with traditional centralized platforms, and supports metadata-driven governance, dynamic visualization, and stateless query execution, thereby allowing equitable and efficient data utilization across enterprise roles, applications, and infrastructure layers.
[0005] Conventional data platforms depend heavily on ingesting and replicating data into centralized memory before enabling data operations. This approach introduces significant latency, consumes storage resources, and complicates compliance with data governance policies. Data operation by ingesting the data also increases the risk of data duplication and inconsistency across distributed systems. The absence of a method for virtual representation of query results executed directly by the processor without storing data limits the ability to perform efficient, real-time data operations across heterogeneous environments.
[0006] Existing data platforms such as FINBOURNE offer automated ingestion and data virtualization capabilities; however, they continue to rely on centralized data preparation workflows and ingestion pipelines, which introduce latency and storage redundancy, particularly in scenarios involving high-frequency or ephemeral data operations. These platforms provide centralized data management systems with integration support via APIs and SDKs, but lack a decentralized marketplace architecture that enables collaborative publishing, rating, and interaction with data products through a dynamic, card-based user interface.
[0007] Furthermore, such systems are primarily cloud-based and abstract hardware-level execution, without optimization for processor-memory interactions or memory-based query execution, thereby limiting their effectiveness in edge computing or resource-constrained environments. While graphical user interfaces and dashboarding tools are available, they do not incorporate fine-grained, real-time visualization mechanisms or adaptive user interaction models, such as status-aware cards and dynamic layout adjustments based on user’s behavior and data state.
[0008] The system disclosed in US10620923B2 provides a centralized data management architecture that supports data ingestion, transformation, and access through structured APIs. However, the architecture relies on persistent data storage and ingestion workflows, which introduce latency and increase infrastructure complexity, particularly in scenarios requiring real-time or transient data access. The system does not implement a unified virtual interface layer capable of establishing direct, transient connections to heterogeneous data sources without replication. Furthermore, the disclosed invention lacks a decentralized data marketplace framework that enables collaborative publishing, adaptive packaging, and dynamic visualization of data products. Hardware-level execution and memory-optimized query processing are not addressed, limiting the system’s applicability in distributed or edge computing environments. The absence of real-time metadata interaction, status-aware visualization, and personalized access control mechanisms further restricts the scalability and responsiveness of the platform in enterprise data democratization contexts.
[0009] US20180301564A1 describes a system for virtualizing access to data across multiple sources. While the system addresses the integration of heterogeneous data, the system still relies on ingestion and replication into a virtualized layer. The system does not support transient query execution or adaptive data packaging and lacks a collaborative marketplace interface.
[0010] US20190046608A1 describes a data marketplace for sharing and monetizing data assets. However, the data marketplace focuses on commercial exchange and lacks technical depth in real-time data access, transient query processing, and hardware-level optimization. The data marketplace does not address dynamic visualization or personalized user experiences.
[0011] US20190345678A1 describes a system that enables real-time access to data streams but is limited to specific data types and does not support unified access across structured, unstructured, and conversational data. The system also lacks a virtual interface layer and adaptive packaging mechanisms.
[0012] Therefore, there is a need for a data platform, a system or a method which overcomes the problems of prior art.
OBJECTS OF THE INVENTION:
[0013] An object of the present invention is to provide a data platform and a method for generating adaptive data packages as virtual representations of query results using a stateless execution model, thereby eliminating the need for persistent storage, ingestion, or replication of data.
[0014] An object of the present invention is to provide a data platform and a method for establishing real-time, transient connections to heterogeneous data sources through a unified virtual interface layer, enabling seamless access to structured, unstructured, and external data without centralized integration workflows.
[0015] An object of the present invention is to provide a data platform and a method for enabling decentralized data interaction via a data marketplace that validates, publishes, and customizes adaptive data packages based on user-specific access permissions and compliance policies.
[0016] An object of the present invention is to provide a data platform and a method deployable across cloud-hosted, on-premises, and hybrid environments as containerized microservices, supporting dynamic query formats and protocol-level interoperability to facilitate scalable and secure data democratization.
SUMMARY OF THE INVENTION:
[0017] The present invention presents a transformative approach to enterprise data management by introducing a data platform that enables real-time data democratization through a transient data marketplace. The data platform is architected to overcome the limitations of conventional centralized systems that rely heavily on data ingestion and replication. In traditional setups, data from various sources—structured databases, unstructured repositories, cloud storage, and external APIs must be ingested into a centralized memory before any meaningful operations can be performed. The ingestion process introduces latency, increases storage requirements, and complicates compliance with data governance policies.
[0018] The invention eliminates these inefficiencies by implementing a platform that establishes transient, memory-based connections to heterogeneous data sources. The platform enables direct access to live data without replication, ingestion, or persistent storage, thereby supporting stateless and secure data operations across distributed computing environments.
[0019] At the core of the platform is a query processing engine that interprets and executes queries in multiple formats, including SQL, JSON Path, and natural language. Upon receiving a query, the engine interacts with the unified virtual interface layer to retrieve relevant data in real time and constructs an adaptive data package. These packages are ephemeral, metadata-rich representations of query results that do not persist in memory. Instead, they are dynamically generated based on the query context, user role, and access policies. The adaptive data packages include schema mappings, transformation logic, and access pointers to the original data sources, ensuring that data is presented in a normalized and consumable format. This stateless execution model supports secure, lightweight, and context-aware data delivery, making it highly suitable for high-frequency, multi-tenant enterprise scenarios.
[0020] The data marketplace component of the invention serves as a decentralized interface for managing and interacting with adaptive data packages. The platform includes modules for approval and publishing, which validate each package against predefined compliance rules and access control policies before making them available to users. The marketplace supports dynamic visualization and personalized user experiences through web portals, APIs, and command-line tools. Users can browse, filter, and retrieve data packages based on their roles and permissions, enabling equitable and efficient data utilization. The platform is deployable across cloud, on-premises, and hybrid environments using containerized microservices, ensuring scalability and fault tolerance. Applications span across industries such as finance, healthcare, logistics, and research, where real-time insights and secure data access are critical. By eliminating ingestion workflows and enabling direct, transient access to live data, the invention redefines the paradigm of data democratization, offering a robust, compliant, and responsive solution for modern enterprise data ecosystems.
BRIEF DESCRIPTION OF DRAWINGS:
[0021] Figure 1 shows a block diagram of a data platform provided in accordance with the present invention; and
[0022] Figure 2, 3, 4 show various detailed schematic views of various components of the data platform shown in Figure 1; and
[0023] Figure 5 shows a flow chart of a method provided in accordance with the present invention.
DETAILED DESCRIPTION OF DRAWINGS:
[0024] In a preferred embodiment of the present invention (Figure 1) a data platform (200) is provided. The data platform (200) is implemented on a network-enabled computing device (100) comprising a processor (110) and a memory (120). The data platform (200) includes a unified virtual interface layer (230), a data marketplace (600), and a query processing engine (240). The network-enabled computing device (100) includes the processor (110), the memory (120), and a network interface (121). The processor (110) executes instructions and manages orchestration logic, while the memory (120) stores platform modules, configurations, and runtime data.
[0025] The network interface (121) enables digital communication with external systems (not shown) using technologies such as Ethernet, Wi-Fi, or cellular modules. The processor (110) executes orchestration logic and task scheduling, while the memory (120) stores platform modules and runtime configurations. The network-enabled computing device (100) is operable in a cloud-hosted infrastructure (100a) or an on-premises computing environment (100b) or a hybrid environment (100c) comprising both cloud and on-premises components.
[0026] In one embodiment (Figure 2), the query processing engine (240) is stored in the memory (120). The QPE (240) is configured to receive a query (q) and generate an adaptive data package (500) as a virtual representation of the query result without storing or ingesting data in a second memory (122). The Second memory (122) can be any type of memory which can be memory of the computing device (100) or a flash memory or any such memory unit. The query processing engine (240) is configured to receive structured or unstructured queries through a data marketplace interface (610) of the data marketplace (600) or system components and initiate real-time data retrieval operations. Upon receiving the query (q), the query processing engine (240) interacts with the unified virtual interface layer (230) to access live data from heterogeneous sources connected therewith without ingesting or storing the data in the second memory (122).
[0027] The query processing engine (240) then dynamically constructs an adaptive data package (500) as a transient virtual representation of the query result, applying schema alignment, filtering, and transformation logic. The adaptive data package (500) is generated on-the-fly and delivered to the data marketplace interface (610) for user access, ensuring lightweight, secure, and context-aware data delivery.
[0028] The adaptive data package(ADP) (500) refers to a transient, virtual representation of query results generated by the query processing engine (240) in response to user or system-initiated queries. The ADP (500) is a dynamically constructed data structure that adapts to the query context, user role, access policies, and real-time data availability. The adaptive data package (500) includes metadata, schema mappings, filtered content, and transformation logic necessary to present the data in a normalized and consumable format. The ADP (500) is generated on-the-fly by accessing live data through the unified virtual interface layer (230), without persisting the data in the memory (120), ensuring lightweight and secure data handling. The ADP(s) (500) are used within the data marketplace (600) to facilitate real-time data interactions, enabling users to view, filter, and retrieve relevant data through various user channels (400, 400a) such as web interfaces, APIs, or command-line tools. The APD (500) can be a data product or a data package or a data set.
[0029] Each ADP (500) is configured to present in a customized view based on a respective user's access permission. The data platform (200) integrates access control logic within the data marketplace (600) and the data marketplace interface (610). Upon generation of an adaptive data package (500) by the query processing engine (240), the data package (500) is tagged with metadata that includes schema definitions, query context, and access attributes. These attributes are evaluated by the approval module (250) before publishing.
[0030] The approval module (250) applies predefined access control policies stored in the memory (120), which are mapped to user roles and permissions. The publishing module (260) then registers the adaptive data package (500) in the data marketplace (600), associating it with a unique query (q1) and access metadata. When a user accesses the data marketplace interface (610), the processor (110) dynamically renders the adaptive data package (500) according to the user's access level. For example, a user with limited permissions may view only summary fields, while a user with analytical privileges may access full schema-transformed datasets.
[0031] The customized view is enforced across user channels (400), including web portals, APIs, and command-line interfaces. The platform (200) presents each adaptive data package (500) in a secure, role-specific format, maintaining compliance and data governance standards. This dynamic rendering mechanism supports real-time personalization without altering the underlying data structure or requiring persistent storage.
[0032] In one embodiment (Figure 2), the query processing engine (240) supports multiple query formats including SQL, JSON Path, and natural language queries, enabling flexible and user-friendly data access across structured and semi-structured sources. Upon receiving a query (q), the engine interacts with the unified virtual interface layer (230) to retrieve relevant data in real time and constructs an adaptive data package (500) without persisting data in memory (120). The adaptive data package (500) comprises metadata describing the query context, schema definitions for structural alignment, and access pointers that reference the original data sources (300). The ADP (500) is generated using a stateless execution model, ensuring that no session-specific data is retained beyond the query lifecycle. The ADP (500) is ephemeral by design, automatically expiring after a predefined time interval or upon completion of the user session, thereby supporting secure and transient data delivery.
[0033] In one embodiment (Figure 3), the query (q) received by the query processing engine (240) originates from a source selected from a group comprising multiple internal and external entities. These sources include a human user (not shown) interacting through the data marketplace interface (610), where queries are submitted via graphical or command-line inputs. Additionally, the query (q) may be generated by an AI module (612) performing predictive or analytical tasks or arise from a service input (613) associated with an ongoing computing activity such as orchestration or monitoring. Queries (q) may also be triggered by service outputs (614) from other computing activities, including CI/CD pipelines or diagnostic modules. Furthermore, external systems (615) may transmit queries via APIs or service connectors, enabling integration with third-party platforms and enterprise applications. Multi-source query architecture supports dynamic, context-aware data access across distributed environments.
[0034] The query processing engine (240) interprets and executes structured queries received from automation workflows or external interfaces, applying transformation, filtering, and aggregation operations on live data streams.
[0035] The data marketplace (600) serves as a modular interface for accessing curated datasets, APIs, and third-party data services, supporting dynamic data retrieval and integration. These components operate in coordination to facilitate real-time, policy-driven data access and processing within the automation platform.
[0036] The data marketplace (600) (Figure 4) is executed by the processor (110) and stored in the memory (120) of the network-enabled computing device (100), forming a centralized environment for managing and interacting with adaptive data packages (500). The data marketplace (600) is operatively coupled to the query processing engine (240). The data marketplace (600) lists generated ADPs (500, 501, and 502) through a display (103) of the computing device (100). Each listed ADP (500) is mapped to a unique query (q1). The data marketplace (600) is virtually connected to the heterogeneous data sources (301, 302).
[0037] The data marketplace (600) includes software modules for cataloguing, indexing, and retrieving data packages, which are dynamically generated, based on query inputs and contextual parameters. The data marketplace interface (610), also stored in the memory (120) and executed by the processor (110), serves as the primary access point for users through one or more user channels (400), such as web portals, APIs, or command-line interfaces. This data marketplace interface (610) is configured to receive structured and unstructured queries, interpret user intent, and facilitate diverse data interactions including browsing, filtering, and accessing data packages. The processor (110) handles execution of query logic and interaction workflows, while the memory (120) maintains session states, metadata registries, and access control policies to support secure and efficient data operations.
[0038] In one embodiment (Figure 1), the data marketplace (600) includes an approval module (250) and a publishing module (260), both instantiated in the memory (120) and executed by the processor (110) of the network-enabled computing device (100). The approval module (250) is configured to validate each adaptive data package (500) against a set of predefined compliance rules, access control policies, and data quality criteria, using rule engines and policy evaluators integrated within the software stack. The approval module (250) is visible through an interface (not shown) of a computing device (any computing device) connected with the data platform (200). A user (validator) can validate through the interface of the data platform (200). Upon successful validation, the publishing module (260) dynamically publishes the adaptive data package (500) to the data marketplace interface (610), enabling real-time access by users. The adaptive data package (500) is not stored in the memory (120). Instead, the ADP (500) is made available as a transient virtual representation, ensuring lightweight and on-demand data delivery without persistent storage.
[0039] The unified virtual interface layer (230) is configured within the memory (120) and executed by the processor (110). The UVIL (230) is operationally coupled to the data marketplace (600), the query processing engine (240) (QPE) and the heterogenous data sources (301, 302). The UVIL (230) accesses data in real-time from the heterogeneous data sources (301, 302) without ingesting or replicating the data into the second memory (122). The unified virtual interface layer (230) includes components such as data connectors, query processors, transformation engines, and communication handlers. The data connectors facilitate protocol-specific access to external sources including databases, APIs, and file systems.
[0040] The query processors interpret and execute data access requests in real time, while transformation engines apply formatting, filtering, or enrichment operations on the retrieved data. The communication handlers manage secure and asynchronous data exchange between the platform and external systems, ensuring consistent and scalable interaction across distributed environments.
[0041] The unified virtual interface layer (230) is implemented as a set of containerized software components deployed within a distributed computing environment, leveraging orchestration platforms such as Kubernetes or Docker Swarm for scalability and fault tolerance. THE UVIL (230) includes protocol handlers for REST, JDBC, and GraphQL to facilitate standardized access to heterogeneous data sources (301, 302).
[0042] The data sources (300) (Figure 1) include a diverse set of repositories including structured databases (300a, 300b), unstructured data repositories (not shown), cloud-based storage systems (not shown), and external APIs. The structured databases (300a, 300b) may include relational systems such as MySQL, PostgreSQL, or Oracle (trade names), supporting SQL-based queries and schema-defined data models. The unstructured data repositories encompass file systems, document stores, and data lakes that store logs, media, or free-form text without predefined schemas. The Cloud-based storage systems refer to services like Amazon S3, Google Cloud Storage, or Azure Blob Storage, which provide scalable object storage accessible via network protocols. The external APIs include third-party services and data providers offering real-time or batch data access through REST, GraphQL, or other standardized interfaces. These heterogeneous sources (301. 302) are accessed by the unified virtual interface layer (230) using protocol-specific adapters and transformation logic to enable seamless data integration.
[0043] The UVIL (230) incorporates schema mapping modules that use metadata registries and transformation engines to align disparate data structures into a unified schema. The data transformation is performed using rule-based processors and mapping templates stored in the memory (120), enabling real-time normalization of incoming data. The mapping modules interact with the network interface to establish virtual connections to external data sources (300), enabling real-time data access. The hardware components (110, 100, and 120) provide computational resources and connectivity, while the software stack orchestrates data normalization and protocol-specific communication across distributed environments.
[0044] In one embodiment, the data platform (200) is deployed as a Software-as-a-Service (SaaS) solution, hosted in a cloud environment such as Amazon Web Services (AWS), Microsoft Azure, or Google Cloud Platform (GCP). The platform (200) is configured to support multi-tenant architecture, allowing multiple organizations or user groups to access the shared data marketplace (600) while maintaining strict isolation of data, configurations, and access controls. Each tenant can independently interact with the data marketplace interface (610), submit queries, and receive adaptive data packages (500) in real time. This SaaS-based configuration ensures scalability, high availability, and ease of integration with enterprise systems, making it suitable for use cases such as data analytics, reporting, and operational intelligence across sectors like finance, healthcare, and logistics.
[0045] The invention eliminates the need for data ingestion through a combination of architectural and functional elements. The query processing engine (240) generates the adaptive data package (500) as a virtual representation of the query result, explicitly without storing or ingesting data into the secondary memory (122). This is supported by the Unified Virtual Interface Layer (230), which establishes real-time, transient connections to heterogeneous data sources (301, 302) and accesses data directly without replication. The Data Marketplace (600) interacts with these components by listing ADPs mapped to queries, not to ingested datasets, thereby maintaining a decentralized and stateless data interaction model. Together, these elements (QPE (240), the ADP (500), the UVIL (230)) enforce a non-ingestive architecture, enabling secure, scalable, and compliant data democratization across distributed environments.
[0046] In one more embodiment of the present invention (Figure 5) a method (1000) for processing and publishing query-based data packages in the data platform (200) is provided.
[0047] The method (1000) starts at step 1000a.
[0048] At step 1001, the unified virtual interface layer (230), the data marketplace (600), and the query processing engine (240) are configured within the memory (120) of the network-enabled computing device (100). This configuration involves instantiating each component as a software module stored in memory (120) and executed by the processor (110).
[0049] At step 1002, the query (q) is received at the query processing engine (240). The query (q,q1) may originate from various sources including user input via the data marketplace interface (610), automated service calls, AI modules, or external systems. The query processing engine (240) is designed to parse and interpret multiple query formats such as SQL, JSON Path, and natural language, enabling flexible and dynamic query intake.
[0050] At step 1003, the query processing engine (240) generates the adaptive data package (500) as the virtual representation of the query result. This generation process involves executing the query logic without retrieving or storing actual data in the secondary memory (122).
[0051] At step 1004, the adaptive data package (500) is validated using the approval module (250) integrated within the data marketplace (600). The approval module (250) applies predefined compliance rules, access control policies, and data quality criteria to ensure that the adaptive data package (500) meets governance standards. Validation is performed in-memory and does not involve data ingestion or replication.
[0052] At step 1005, the validated adaptive data package (500) is published to the data marketplace interface (610) via the publishing module (260). The publishing module (260) registers the adaptive data package within the marketplace (600), making it accessible to authorized users and systems. The publication process includes tagging the package with metadata and associating it with its originating query.
[0053] At step 1006, the adaptive data package (500) is listed in the data marketplace (600), where each package is mapped to the unique query (q1). The listing mechanism enables indexing and retrieval of adaptive data packages based on query identifiers, metadata attributes, and user access permissions, facilitating organized and efficient data interaction.
[0054] At step 1007, the view of each listed adaptive data package (500) is customized for individual users based on their access permissions. The data marketplace interface (610) dynamically adjusts the presentation of the adaptive data package, ensuring that users only see data elements they are authorized to access. This customization is enforced through access control policies embedded in the approval module.
[0055] At step 1008, the query processing engine (240) is operatively coupled to both the unified virtual interface layer (230) and the data marketplace interface (610). This coupling enables seamless communication between query execution, data access, and user interaction components. The query processing engine (240) utilizes the unified virtual interface layer (230) to access data sources and the data marketplace (600) interface to deliver adaptive data packages (500).
[0056] At step 1009, virtual connections are established between the data marketplace (600) and the plurality of heterogeneous data sources (301, 302). These connections are facilitated by the unified virtual interface layer (230), which supports multiple data access protocols and performs schema normalization. The connections are transient and do not involve data replication or ingestion.
[0057] At step 1010, data is accessed in real-time from the heterogeneous data sources (301, 302) via the unified virtual interface layer (230), without ingesting or replicating the data into the second memory (122). The unified virtual interface layer (230) executes queries directly against the source systems, retrieving only the necessary pointers and metadata to construct the adaptive data package, thereby maintaining a stateless and non-ingestive data access model.
[0058] The method (1000) ends at step 1000b.
[0059] Users can utilize the data platform (200) to access and interact with real-time data from multiple sources (301, 302) without needing to move or store the data. By submitting a query (q) through the query processing engine (240), users receive an adaptive data package (500) that represents the result virtually. This is especially useful for analysts, researchers, or business users who need instant access to insights across distributed systems, such as querying sales data from different vendors or accessing patient records across hospitals.
[0060] Through the method (1000), users can validate and publish these adaptive data packages (500) via the data marketplace (600), making them accessible to others based on permissions. Users can customize views (1007) of the data packages depending on their roles such as a compliance officer reviewing only validated financial data or a data scientist accessing schema-transformed datasets for modeling. The platform (200) supports various query formats (SQL, JSON Path, natural language), allowing users from technical and non-technical backgrounds to interact with data seamlessly.
[0061] The present invention has an advantage of advancing the data democratization by enabling secure, scalable, and real-time access to diverse data sources without centralized ingestion or replication. The data platform (200) achieves this through the unified virtual interface layer (230), which establishes transient, memory-based connections to the heterogeneous data sources (301, 302) and allowing users to retrieve data dynamically without storing it in a second memory (122). This eliminates the latency, duplication, and compliance risks associated with traditional ingestion-based systems. The query processing engine (240) generates the adaptive data packages (500) as ephemeral, stateless representations of query results, ensuring that data operations are performed directly in the memory (120) and the processor (110), aligned with distributed computing environments.
[0062] The method (1000) reinforces this architecture by defining steps for configuring the platform components (1001), executing queries (1002), and publishing validated data packages (1005) in a decentralized data marketplace (600). Users can interact with these packages in real-time, with customized views (1007) based on access permissions, supporting equitable and efficient data utilization across enterprise roles. By eliminating ingestion workflows and enabling direct, transient access to live data, the invention overcomes the limitations of centralized platforms and supports responsive, metadata-driven governance and visualization core requirements for true data democratization. , Claims:We Claim:
1) A data platform (200) implemented on a network-enabled computing device (100) comprising a processor (110) and a memory (120), the data platform (200) comprising:
a query processing engine (240) (QPE) stored in the memory (120), the QPE (240) is configured to receive a query (q) and generate an adaptive data package (500) (ADP) as a virtual representation of the query result without storing or ingesting data in a second memory (122);
a data marketplace (600) executed by the processor (110) and stored in the memory (120), the data marketplace (600) is operatively coupled to the query Processing Engine (240), wherein the data marketplace (600) lists ADP (500) and each listed ADP (500) is mapped to a unique query (q1), and the data marketplace (600) is virtually connected to the heterogeneous data sources (301, 302 );
a unified virtual interface layer (230) (UVIL) configured in the memory (120) and executed by the processor (110); the UVIL (230) is operationally coupled to the data marketplace (600), the query processing engine (240) (QPE) and the heterogenous data sources (301, 302), wherein the UVIL (230) accesses data in real-time from the heterogeneous data sources (301, 302) without ingesting or replicating the data into a second memory (122).
2) The data platform (200) as claimed in claim 1, wherein the unified virtual interface layer (230), the query processing engine (240), and the data marketplace (600) are each configured to be instantiated and executed within the memory (120) of the network-enabled computing device (100), and the network-enabled computing device (100) is operable in one or more deployment environments selected from the group consisting of a cloud-hosted infrastructure (100a), an on-premise computing environment (100b), and a hybrid environment (100c) comprising both cloud and on-premise components.
3) The data platform (200) as claimed in claim 1, wherein the data marketplace (600) comprises an approval module (250) configured to validate the adaptive data package (500) based on predefined compliance rules, access control policies, and data quality criteria, and a publishing module (260) configured to publish the validated adaptive data package (500) to the data marketplace (600) and accessible through a data marketplace interface (610) of the data marketplace (600).
4) The data platform (200) as claimed in claim 1, wherein the unified virtual interface layer (230) is implemented as a containerized microservice deployed in a distributed computing environment, supports data access protocols and is configured to perform schema mapping and data transformation to normalize heterogeneous data formats from the data sources (301, 302).
5) The data platform (200) as claimed in claim 1, wherein each ADP (500) is configured to present in a customized view based on a respective user's access permission.
6) The data platform (200) as claimed in claim 1, wherein the query processing engine (240) supports one more query formats and natural language queries; wherein the adaptive data package (500) comprises metadata, schema definitions, and access pointers to relevant data sources (300), is generated using a stateless execution model, and is ephemeral, expiring after a predefined time interval or upon completion of the user session.
7) The data platform (200) as claimed in claim 1, wherein the query received by the query processing engine (240) originates from a source selected from the group consisting of: a human user via the data marketplace interface (610), an AI module (612), a service input (613) from a computing activity, a service output (614) from another computing activity, or an external system (615).
8) The data platform (200) as claimed in claim 1, wherein the data platform (200) is deployed as a Software-as-a-Service (SaaS) solution, accessible to users via a cloud-hosted environment, and configured to provide multi-tenant access to the data marketplace (600) through user channels (400).
9) A method (1000) for processing and publishing query-based data packages in a data platform (200) implemented on a network-enabled computing device (100), the method (1000) comprising steps of:
configuring (1001) a unified virtual interface layer (230), a data marketplace (600), and a query processing engine (240) within a memory (120) of the network-enabled computing device (100);
receiving (1002) a query (q) at the query processing engine (240);
generating (1003) an adaptive data package (500) by the query processing engine (240) as a virtual representation of the query result;
validating (1004) the adaptive data package (500) using an approval module (250);
publishing (1005) the validated adaptive data package (500) to a data marketplace interface (610) via a publishing module (260);
listing (1006) the adaptive data package (500) in the data marketplace (600), wherein each adaptive data package (500) is mapped to a unique query (q1);
customizing (1007) the view of listed adaptive data packages (500) for each user based on their access permissions;
coupling (1008) the query processing engine (240) operatively to both the unified virtual interface layer (230) and the data marketplace interface (610);
establishing (1009) virtual connections between the data marketplace (600) and a plurality of heterogeneous data sources (301, 302); and
accessing (1010) data in real-time from the heterogeneous data sources (301, 302) via the unified virtual interface layer (230), without ingesting or replicating the data into a second memory (122).