Backdocket makes AI a first-class citizen. The Claude, ChatGPT, or Gemini subscription you already have now works on your cases. No cost increase.
Don’t pay the markup.
Claude, ChatGPT, and Gemini are the frontier models. Anthropic, OpenAI, and Google spend billions of dollars making them the best AI in the world — and your firm can already use them for the price of a subscription.
Most legal AI products are wrappers. They take one of those same models, put their own screen in front of it, and charge you for the model plus their margin. Then they ask you to copy your case data out of your practice management system and into their product so the AI can see it. You pay twice and maintain your data twice, for an AI you already had.
Backdocket skips the wrapper. Your case data stays where it already lives, and Claude, ChatGPT, or Gemini connects to it directly through a secure, authenticated connection — no copies of your data in someone else’s product, no middleman, no cost increase.
Every Backdocket endpoint is accessible through the Model Context Protocol — an open standard for AI-to-application integration. Firms and integrators can build custom AI-powered workflows, connect external AI tools to their case data, and automate processes programmatically.
Because it’s MCP, the frontier models work with Backdocket directly — Claude, ChatGPT, Gemini, or any other tool that supports MCP connectors, such as Microsoft Copilot. Your team opens the model they already use and asks it to pull up a claim, summarize the medical records, draft a demand, or run a report — and it does, using your live Backdocket data.
These models aren’t limited to answering questions when someone asks. Claude and ChatGPT can run tasks on a schedule, and through MCP they can call the AI agents built into Backdocket — the Case Analyst, Medical Chronologist, Demand Analyst, Billing Auditor, and the rest. Set your assistant to review new documents every morning or check demand readiness every week, and it does the work on its own.
Verified compatible with Zoom AI Companion. Firms already using Zoom AI Companion can add Backdocket as a custom MCP connector and query case data, run reports, or trigger workflows directly from a Zoom meeting or chat — no separate integration project required.
Now on the Microsoft Marketplace: the Backdocket connector for Microsoft 365 Copilot. Ask Copilot Chat or Researcher about your cases in plain language. Find claims, review status and key dates, look up contacts, check open tasks and deadlines, and run your saved Backdocket searches. Each user signs in with their own Backdocket account and sees only what they can already see. Your Microsoft 365 admin approves it once for the firm. Read-only by design: Copilot looks, it doesn’t change records.
When an AI vendor resells model access, you pay for the model plus their margin. Why pay someone else to use the model when you can use it yourself and skip the fees? Backdocket is built on a simple principle: value is a core value. So we don’t sit between you and your AI provider collecting a fee.
For firms that want AI inside Backdocket’s own screens, the platform includes a full built-in AI suite that runs on your own model and API key. Everything below — including the specialized sub-agents and background agents — is also available to Claude, ChatGPT, or Gemini through MCP.
A ChatGPT-style conversational interface scoped to individual claims. Ask questions about case documents in plain language and get citation-backed answers drawn from your actual files — not generic training data. Responses stream in real time — token by token — so your team sees answers as they’re generated, not after a loading delay.
Every response includes inline citations linking to the specific source document and page number. Citations aren’t limited to documents — the AI also cites claim notes, ledger entries, contacts, and claim fields, giving your team a complete evidentiary trail for every answer.
A single conversation can reference multiple claims simultaneously — enabling cross-case analysis, pattern identification, and comparative review without switching between threads. Tag specific documents to focus the AI on just those files, or leave it open to search across everything on the claim.
Create, search, rename, and organize conversations. Server-side search lets you find past conversations by title or tagged claim — even claims that were later deactivated.
Firm-wide conversational AI for general legal questions, cross-case analysis, and operational queries. Not scoped to a single claim — use it for firm-wide insights, comparative analysis, and general research.
Save any AI response or generated document directly to the claim record as a permanent document — one click. When multiple claims are tagged, choose which claim to save to. Generated files are available for immediate download and remain accessible for 30 days; saved copies are permanent.
When you tag a claim for AI analysis, Backdocket automatically ingests every document — extracting text, chunking it using LLM-assisted semantic segmentation, generating embeddings, and indexing for vector search. New documents added to the claim are automatically indexed without manual intervention. Each firm gets its own dedicated vector index — structural tenant isolation, not shared infrastructure with query-level filtering.
When you ask a question, the system retrieves the most relevant document segments, reranks them for accuracy, and provides the AI with your actual case records as context. The result: answers grounded in your files, not hallucinated from training data. If vector search is temporarily unavailable, the AI continues working in degraded mode — graceful fallback, never a hard failure.
The embedding and vectorization infrastructure is powered by Amazon Bedrock and paid for by Backdocket — not by your firm. The rest of the ingestion pipeline — OCR, LLM-assisted chunking, and entity extraction — runs on your model with your key, like every other AI capability in the platform.
Built-in OCR automatically extracts text from scanned medical records, bills, and images. The AI reads scanned documents just as accurately as digital files — no external service, no manual preprocessing. Like every AI capability in Backdocket, OCR runs on your model with your key.
Dates, dollar amounts, and named parties are automatically extracted from case documents at upload time. Extracted entities are stored in a structured, queryable format — enabling instant medical chronologies, expense summaries, and party rosters without manual review.
Complete document text is stored and retrievable through a three-tier access model: entity lookup for structured data, semantic vector search for targeted questions, and full document text retrieval when the AI needs exhaustive access. Your AI always has the complete picture.
Generate professional documents directly from AI conversations — demand letters, settlement brochures, billing summaries, and correspondence as downloadable Word, PDF, or Excel files populated with real case data.
A capability unique to Backdocket. Three drafting modes allow firms to leverage their existing templates with AI:
No other legal practice management platform offers three-mode template drafting. Your templates, your formatting, your data — assembled by AI.
AI fills your existing Word merge templates with real case data while preserving original formatting. Prompt resolution supports every merge field type in the system — Claim, ClaimContact, Staff, FillIn, Claimant, Request, Ledger, and Contact. Professional documents that look exactly like your firm’s templates.
The AI doesn’t just chat — it deploys domain-expert sub-agents for specific legal tasks:
Each sub-agent receives focused context and brings specialized expertise to its analysis. The system selects the right sub-agent automatically based on the task. Beyond predefined roles, the AI can create custom sub-agents with firm-specific instructions for specialized workflows unique to your practice.
Generate pie, donut, bar, column, line, area, and scatter charts directly from AI conversations. Ask questions about your case data and get visual answers — multiple charts per message, automatic data normalization for small categories, and interactive tooltips. Turn case data into actionable insights without leaving the conversation.
AI that works without being asked. Background agents are triggered automatically by case events — document uploads, date changes, elapsed time — not by user messages.
No competitor has shipped this breadth of proactive AI automation. This is the most comprehensive AI automation suite in the plaintiff software market — built into the platform and running on your own model and key.
Choose your AI provider — Anthropic (Claude), OpenAI (GPT), or Azure OpenAI. Each firm configures their own provider and API key. Switch providers anytime with no data migration, no re-indexing, and no disruption.
No other legal practice management platform offers provider choice. Every competitor locks you into their AI vendor.
Every major LLM provider supports multilingual translation natively. Backdocket’s AI can translate documents, correspondence, and case notes to and from client languages directly within the platform — no external translation service or add-on required.
Monitor AI token usage by provider, model, and call type. Monthly usage summaries on the admin dashboard with 24-month retention. A visual context utilization fill bar displays in real time during conversations — your team always knows how much capacity remains. Cancel any in-progress AI operation at any point if the response isn’t going in the right direction.
Claude, ChatGPT, and Gemini are the world’s best AI. Backdocket gives them first-class access to your cases through MCP, with no cost increase — and, for firms that want it, a full built-in AI suite with citation-backed answers, document generation, template-aware drafting, specialized sub-agents, proactive background automation, and AI charting. Use the real AI. Choose your provider. Control your costs.
$59.99/user/month. Don’t pay the markup. Use Claude, ChatGPT, or Gemini directly through MCP, or bring your own model and key for the built-in AI — your provider bills you directly, with no markup from Backdocket. No vendor lock-in.
View full feature list
