What it does

View and inspect

• Documents view. Every record in the database, in a sortable grid you can filter and sample.
• Conversation view. See messages grouped into the conversations they belong to, with participants, dates, and message counts.
• Native preview. Render attachments — PDFs, images, office files — without leaving the application.

Clean and prepare

• Deduplicate records using the standard Bates approach or a custom hash on fields you choose.
• Name normalize contacts so John Smith, j.smith, and Smith, John collapse into one identity. Original names are preserved so you can roll back any decision.
• Time-zone convert message dates between source and target zones — including a "process in a different zone than the load file" workflow for multi-jurisdictional reviews.
• Project-level conversations treat the same Slack channel imported from two different exports as one conversation, so participants and messages combine cleanly.

Search and tag

• Search bar with default and custom saved searches, AND/OR logic, and a Names-search mode for finding everything someone said.
• dtSearch keyword indexing for full-text search across message bodies and extracted text. Supports proximity, phrase, fuzzy, and wildcard syntax.
• Tagging with five modes: tag the document, the family, the conversation, the day, or a configurable range of messages around it.

Transform and export

• Mass update, search and replace, date format, and renumber tools for bulk changes across a search.
• Export to RSMF, PDF, EML, HTML, or DAT — or all of them at once via Universal Export.

Data sources supported

Message Crawler reads chat and short-message data from a wide variety of sources:

• Chat platforms. Slack, Microsoft Teams (multiple HTML versions plus PST), WhatsApp, Telegram, Discord, Zoom, Webex, MatterMost, Zoho Cliq, Lark, Meta Workspace, Facebook, Instagram, LinkedIn, ChatGPT.
• Forensic and mobile-collection tools. Cellebrite, Oxygen, Axiom, Elcomsoft, ModeOne, Belkasoft, XRY, iMazing, SyncTech, Apple Chat.DB.
• Trading-floor and email tools. Bloomberg IM, Bloomberg XML, Google MBOX/PST, Google JSON, MBOX (including Google Vault and Android SMS sync), Threema, Twitter.
• Direct-format imports. RSMF (with built-in fixer for malformed files), JSON, XML, SQLite, ZIP archives.

If your data is in a format not on this list, ask. Adding new sources is part of the regular development cycle.

Export formats

• RSMF (v1 and v2). Relativity Server 2023 and earlier uses v1; RelativityOne and Server 2024 use v2.
• PDF — including embedded native attachments for self-contained delivery, with optional Bates stamping.
• EML — for review platforms that ingest email-style chat.
• HTML — for direct sharing or web-style preview.
• DAT — Concordance-compatible load files.
• Universal Export — produce multiple formats simultaneously from a single job.

How it works

Message Crawler is built around four objects: Documents (everything loaded), Searches (saved queries that drive most other work), Jobs (how import, transform, and export actually happen), and Log (everything that's been done).

Jobs use a distributed model — large datasets break into smaller tasks that run in parallel across as many agents as you make available. A real-world Slack import of 32 million records on a typical SQL Server has been processed in about six hours.

You choose your database backend at setup: MS SQL for production work, MySQL as a free alternative, SQLite for small datasets and demos, or a Custom connection string for unusual setups.