1. Latest Release Home
  2. Guides
  3. Concepts
    1. Aggregate filter
    2. Alias
    3. Apache Avro™
    4. Apache Kafka™
    5. Applied filter
    6. Base table
    7. Bookmark
    8. Business schema
    9. Business schema view
    10. Business schema view column
    11. Business schema view formula column
    12. Column
    13. Connector
    14. Dashboard filter
    15. Dashboard folder
    16. Dashboard runtime filter
    17. Data agent
    18. Data alert
    19. Data destination
    20. Data purge
    21. Data retention
    22. Date system variable
    23. Direct query
    24. Distinct filter
    25. Dynamic field
    26. Expression
    27. External data source
    28. External session variable
    29. External table
    30. Filter expression
    31. Filter expression session variable
    32. Filter option
    33. Formula column
    34. Formula expression
    35. Full load
    36. Global variable
    37. Incorta Analyzer Table
    38. Incorta Analyzer Table column
    39. Incorta Analyzer Table formula column
    40. Incorta Analyzer View
    41. Incorta Analyzer View formula column
    42. Incorta RAG
    43. Incorta Spark SQL View
    44. Incorta SQL Table
    45. Individual filter
    46. Insight filter
    47. Internal query expression
    48. Internal session variable
    49. Join
    50. Join condition
    51. Join condition column
    52. Join filter
    53. Join filter column
    54. Level-based measure
    55. Load filter
    56. Local data file
    57. Local data folder
    58. Log-based incremental load
    59. Materialized View
    60. Materialized View column
    61. Materialized View formula column
    62. Measure filter
    63. Miscellaneous system variable
    64. Multi-source table
    65. Physical schema
    66. Physical schema table
    67. Physical schema table column
    68. Physical schema table formula column
    69. Physical schema table runtime security filter
    70. PostgreSQL (Incorta SQL) View
    71. Presentation variable
    72. Prompt
    73. Query system variable
    74. Remote table
    75. Runtime business view
    76. Runtime business view formula column
    77. Runtime security filter
    78. Schema evolution
    79. Service Status Observability
    80. Session variable
    81. Shared Storage
    82. Snowflake result set
    83. Snowflake view
    84. Upgrade Readiness
    85. Variable
    86. Visualization
  4. Tools
  5. References
  6. Integrations
  7. Data applications
  8. Data Governance
  9. Release Notes
  10. Incorta Premium
  11. Release Support Policy
  12. Release Model

Concepts → Incorta RAG

Overview

Incorta Retrieval-Augmented Generation (RAG) introduces native unstructured data management capabilities that enable users to upload, process, manage, and use files directly within the platform. Users can create knowledge stores from a dedicated subtab under the Content tab, add files to those stores, and the content becomes available in Incorta Nexus for semantic search and RAG.

Uploaded files are automatically processed and indexed using configured AI services, enabling users to retrieve contextual information and interact with enterprise documents through AI chat experiences.

The Incorta RAG capability streamlines the management of unstructured content and simplifies AI-powered knowledge retrieval workflows within Incorta.

Note

Incorta RAG is available starting 2026.3.0

Important

Incorta RAG requires a premium cluster, Incorta Nexus enabled, and configured AI services, including a Large Language Model (LLM), Embedder, and Reranker. For information on enabling Incorta RAG, refer to Guides → Configure Server.

Knowledge Stores

The Knowledge Stores subtab is available under the Content tab in Analytics. Knowledge stores are repositories for organizing and managing unstructured content within Incorta. Users can create knowledge stores and upload files that are automatically processed for semantic retrieval.

Once a knowledge store is enabled for Incorta Nexus, its content becomes searchable through AI chat experiences.

Knowledge stores simplify the management of enterprise documents by centralizing unstructured content in a format optimized for retrieval and contextual response generation.

Semantic Search and RAG

Incorta RAG uses embedding, reranking, and language model services to support semantic retrieval workflows.

When a user submits a prompt through AI chat:

  1. Relevant document chunks are retrieved from the configured knowledge stores.
  2. Retrieved content is reranked for relevance.
  3. The LLM generates contextual responses using the retrieved information.

This process enables natural language interaction with uploaded enterprise content and improves the relevance of generated responses.

Limitation

AI service configurations are applied at the cluster level and cannot be configured per knowledge store.