Sorana

Sorana is an AI-powered visual organization platform that transforms interaction with digital files. It uses semantic analysis to automatically group related files and folders onto a two-dimensional spatial canvas, replacing list hierarchies with visual arrangements. The platform allows creating workspaces via drag-and-drop, as well as orchestrating no-code multi-agent workflows, connecting to local or cloud AI backends while maintaining control over data.

★★★★★

5.0(1 ratings)

• Spatial Organization with AI. Sorana's main feature uses semantic artificial intelligence algorithms to analyze content, metadata, and contextual relationships between files and folders. Subsequently, it represents these elements on a two-dimensional canvas, visually grouping them according to their thematic or project similarity. This spatial representation replaces the nested directory model with a map view that shows hidden connections and structures in the data.
• Interactive Visual Canvas Editor. This functionality provides a direct editing environment to manipulate elements within the workspace. Users can reorganize entire groups, move individual files between different canvas zones, resize containers, and modify group boundaries through drag-and-drop actions. Changes in the visual arrangement are immediately reflected in the interface.
• Advanced AI Model Management. A centralized system allows configuring and managing connections with multiple large language model backends. It supports cloud services like OpenAI and Mistral, alongside local solutions like Ollama and Llamacpp. The manager lists all available models, allowing users to activate or deactivate specific instances for particular tasks and configure individual connection parameters.
• No-Code Multi-Agent Orchestration. Through a visual interface, users can design complex workflows involving multiple specialized AI agents. These agents are created and configured from the interface, defining their purpose, tools, and assigned AI models. They are then connected in sequences or processing networks by dragging nodes onto the canvas, establishing communication channels between them to pass data and results.
• Knowledge Integration in Agents. AI agents within Sorana can augment their contextual memory and processing capacity through document integration. The function processes PDF files using OCR and analyzes text documents to extract structured and semantic information, which is then incorporated into the agent's operational context, enabling it to perform tasks based on the content of those specific documents.
• Interactive Document Chatbot. It implements a conversation system that allows direct interaction with the content of files uploaded to the workspace. Users can ask questions in natural language about information contained in specific documents or groups of documents, and the system provides precise answers by extracting and synthesizing relevant information from the analyzed content.
• Visual Structure Map Generation. It analyzes folder hierarchies and projects them as dynamic visual representations akin to mind maps or network diagrams. This function reveals organization patterns, dependencies between files, and the overall structure of projects, facilitating the understanding of relationships that are not evident in conventional list views.
• Data Portability and Control. The application is designed as a portable solution that operates while keeping all user data under their direct control. Workspaces, configurations, and metadata generated by the AI are stored locally or in user-specified locations, without relying on centralized servers for the main storage of organizational information.
• Visual Workflow Configuration. It provides a dedicated module to define execution parameters for automation pipelines. Users can adjust properties such as activation conditions, error handling, input parameters for agents, and output formats through configuration panels associated with each node of the flow on the visual canvas.
• Connectivity to Processing Backends. It establishes and manages stable connections with various AI processing services, both local and remote. The function handles authentication, request submission, response reception, and connection error management, providing a unified abstraction layer to interact with different APIs and protocols of AI services.
• Automatic Semantic Content Analysis. It performs continuous processing of file content added to the workspace, extracting entities, topics, key concepts, and contextual relationships. This analysis forms the basis for automatic visual groupings and enables search and filtering functionalities based on meaning, not just file names or basic metadata.
• Manual and Automatic Grouping System. It combines automatic AI organization with manual categorization tools. Users can create new groups by defining custom semantic criteria, adjust automatically generated groups, and reorganize elements between different categories, always maintaining visual coherence on the canvas.

Sorana's development began in response to perceived limitations in traditional file managers. The development team, Tetramatrix, identified that nested list and folder hierarchies do not reflect how human cognition works with spatial associations and relationships. The creation of the project formally started with the goal of building a system that organizes digital information based on meaning and visual context rather than storage locations.

Alternatives to Sorana: