AUM：A Private Enterprise AI Knowledge Base Client Running Locally Offline

AUM：A Private Enterprise AI Knowledge Base Client Running Locally Offline

aumbot.co/

AUM is an AI intelligent engine designed specifically for enterprises, supporting local offline operation to ensure data security and privacy. It helps enterprises efficiently manage and analyze internal documents, supports multiple mainstream open-source models, and meets the customized needs of different departments.

AUM is an intelligent AI engine specially created for enterprises, officially referred to as the “enterprise brain.” It can run entirely locally on users’ own servers or computers without Internet connection, ensuring complete confidentiality and security of enterprise data. Users can interact with AUM in a conversational manner, instructing it to process and analyze various internal company files, reports, and databases to quickly obtain required information, generate documents, or complete data analysis. This architecture effectively avoids data leakage, regulatory violations (such as HIPAA, GDPR), and intellectual property risks potentially caused by cloud-based AI services.

The system supports integration with multiple mainstream open-source AI models, allowing enterprises to choose and deploy as needed, such as installing on fully physically isolated servers. AUM can also create customized AI models for different business teams (e.g., HR or Legal departments) to help them efficiently accomplish their respective tasks, such as quickly screening resumes or analyzing contracts.

Illustration

Features List

  • Fully Local Deployment: Supports installation on internal enterprise servers, private clouds, or completely isolated devices, ensuring data does not leave the enterprise network.
  • Enterprise-grade File Search: Quickly retrieves massive amounts of internal company documents and private materials for precise information pinpointing.
  • Multi-File Collaborative Processing: Capable of processing large batches of documents simultaneously, such as batch resume or contract analysis, completing corresponding tasks based on instructions.
  • Database Conversational Analysis: Allows users to ask databases questions in natural language directly, simplifying data analysis and eliminating the need for complex queries.
  • Customizable AI Brain: Supports building dedicated AI for different departments (e.g., HR, Legal, Operations) tailored to business process needs.
  • Workflow Automation: Can automatically create documents, draft proposals, send emails, or write data into applications like Notion, improving efficiency.
  • Supports Mainstream Open-Source Models: Compatible with models including Llama3, Mistral, Mixtral, DeepSeek, etc., giving enterprises freedom of choice.
  • Permission Management System: Built-in role-based access control supports fine-grained permission management.

Usage Help

AUM is a locally deployed AI tool designed to help enterprises analyze and handle sensitive internal information with AI technology without data leaving the premises.

Installation and Deployment

AUM offers multiple deployment options to meet different enterprise security needs, all without requiring connection to the public Internet:

  1. On-Premise Server Deployment

    • Enterprises prepare servers with sufficient performance.
    • Install the AUM software on these servers.
    • Employees access AUM through a browser pointing to the server address; all data processing occurs inside the firewall.

  2. Private Cloud Deployment (VPC)

    • Enterprises using cloud services like Alibaba Cloud, Tencent Cloud, AWS, Azure, and owning a Virtual Private Cloud (VPC), can deploy AUM there.
    • VPC environments are isolated from the public network, ensuring data security and access permission only to authorized personnel.

  3. Air-Gapped Infrastructure Deployment

    • Suitable for institutions with extremely high security requirements such as government, military, and core financial departments.
    • AUM is installed on completely network-disconnected standalone computing devices.
    • Operators must physically access the device, providing the highest level of security.

  4. Multi-Device Local Deployment

    • Upcoming feature supporting distributed AUM installation on multiple local devices to enable collaborative work.

Core Functional Operations

AUM is primarily used through a conversational interface similar to chatbot interaction:

1. Asking Questions and Searching Files

No need to open files one by one; just enter a question to quickly get answers.

  • Open the AUM main interface.
  • Enter queries in the chat box, e.g., “Help me find all contracts for Project A from last year” or “Which product had the highest sales in the Q3 financial report?”
  • AUM automatically searches relevant files and directly responds with answers.

2. Processing Multiple Files Simultaneously

Departments like HR can bulk filter data.

  • Upload or specify resume files to the AUM workspace.
  • Enter instructions, e.g., “Filter all candidates with over 5 years of Java experience and graduated from 985 universities.”
  • AUM reads and organizes the list matching these criteria. Legal teams can simultaneously compare multiple contract clauses.

3. Conversing with Databases

Marketing managers can access databases without SQL.

  • Technicians first connect the sales database to AUM.
  • Users ask in natural language, e.g., “Which channel brought the most new users last month?” or “Generate sales trend charts for each product over the past six months.”
  • AUM converts questions into query instructions and presents results in charts or text.

4. Creating Customized AI Brains

Build dedicated AI services for specific departments.

  • Admin selects “Create New Brain” in the backend.
  • Upload department-specific materials like employee manuals, recruitment standards, etc.
  • Department staff base questions on these materials to obtain precise and compliant answers, e.g., inquire about onboarding processes or generate job descriptions.

Application Scenarios

  1. Law Firms Lawyers quickly search and analyze contracts, evidence, and legal documents, eliminating risks of client information leaks.

  2. Medical Research Institutions Hospitals deploy internally to assist researchers in compliant analysis of anonymized medical records and latest literature, accelerating medical research.

  3. Financial Institutions Investment banks securely analyze market reports and financial data, ensuring confidential information is not leaked.

  4. Government and Defense Departments Deployed in air-gapped networks to ensure confidential intelligence and policy internal security management.

  5. Large Enterprise Internal Knowledge Base Centralizes various scattered documents, building a unified intelligent knowledge center, helping employees quickly query information.

Q&A

  1. What is the difference between AUM and cloud-based AI like ChatGPT? All data in AUM stays locally within the enterprise, protecting data privacy and control; cloud AI uploads data, posing leakage and compliance risks.

  2. What are the hardware requirements for running AUM? Supports high-performance hardware; stronger resources can unlock more advanced AI models and faster responses.

  3. Which AI models does AUM support? Supports multiple mainstream open-source large language models such as Mistral, DeepSeek, Llama3, Mixtral, etc., with flexible choices.

  4. Is there a free version of AUM? Yes, the official website offers a permanently free version with no access restrictions.

Language:
EN ZH