Skip to content

The Documentation is a Work in Progress

🚧🚧🚧🚧🚧 We are still writing the documentation prior to an official release 🚧🚧🚧🚧

Archivist is an easy-to-use desktop application that:

1. Brings local, offline, AI chat capabilities to your personal computer; and
2. Enables you to safely utilize your own personal/private/proprietary files and documents for the AI to reference

It is designed for people who want to leverage the power of AI without relying on cloud providers.

What makes Archivist special is its ability to learn from your own files and documents. Simply upload any files you want—like PDFs, Word documents, or text files—and Archivist makes their content available during your AI conversations. You can ask questions about specific documents or across your entire collection, getting insightful answers based on your personal information. No technical knowledge required—just upload your files, ask questions, and get smart responses that draw directly from your own curated library of documents.

About the Name

An archivist is an information professional who assesses, collects, organizes, preserves, maintains control over, and provides access to records and archives determined to have long-term value... Archivists keep records that have enduring value as reliable memories of the past, and they help people find and understand the information they need in those records. Source: Wikipedia

This application serves as a capability demonstration rather than a finished product. As system integrators specializing in AI solutions for professional services firms, developing mass-market desktop applications isn't our core business model.

We've made deliberate packaging decisions that prioritize demonstrability over startup speed. The application wraps what would typically be an always-on network service into a distributable format you can try on your own hardware. There is an inherent speed tradeoff with this approach.

Query Tab
Pre-Process Tab
Upload Tab
Browse Tab
Inspect Tab
AI Settings Tab
Help and Licensing Tab

Value Proposition

  • Local/private/secure [[Retrieval Augmented Generation|RAG]] application for individual users
  • Fully functional at no cost
  • No subscription fees
  • No service outages
  • No chat limits
  • Easy setup
  • Low cost one-time paid [[Licensing]] option to enable users to upload custom LLMs
  • Support [[Open Source]]
  • Powered by commercial-grade AI technology by [[IBM]]
  • Makes your data [[Data Portability|portable]] for other AI applications
  • Runs on a laptop

  • Swap models as technology advances

  • Cost advantage over time

    • No ongoing subscription fees to ChatGPT Plus ($20/month), Claude Pro ($20/month), etc.
    • One-time payment vs. potentially hundreds of dollars per year
    • Clear ROI after just a few months compared to commercial AI subscriptions

  • Reliability and independence

    • Works offline, including in areas with poor internet
    • Not subject to API rate limits or service outages
    • No need to worry about price increases

  • Resource efficiency

    • Optimized for their specific use case
    • Better performance on targeted tasks than general-purpose AI
    • Faster responses for document-specific questions

Feature Overview

  • Your personal, purpose-built, [[Retrieval Augmented Generation]] application

  • Uses metadata filtering to enable you to precisely control which information the AI considers—whether you want answers from a specific individual document, a curated collection of related files ([[File Sets]]), or your entire knowledge base—ensuring your conversations remain focused on exactly the information that matters to your current task.

  • Simple AI chat with a local LLM

  • Voice input so you can dictate your messages

  • Pre-processing text

  • Infinitely flexible text chunking strategies

  • Simple import/export to improve [[Data Portability]] and reduce vendor lock-in

Explore the power of private data intelligence with our showcase application! This demonstration highlights our expertise in:

  • Seamless ingestion of your proprietary documents
  • Intelligent chunking and vector embedding
  • Optimized retrieval with contextual awareness
  • Natural language generation grounded in your data
  • Secure, local processing for sensitive information