AI Agents Overview

CaseSolo includes AI-powered agents that automate analysis, drafting, and research tasks.

What Are AI Agents?

AI agents are specialized tools that analyze your case data and produce actionable outputs. Each agent is designed for a specific task — from summarizing medical records to drafting demand letters.

Available Agents

Intake Qualifier

Analyzes a new case intake to assess viability, liability strength, and estimated value.

• Input: Case details, incident description
• Output: Qualification score, risk factors, recommended next steps

Demand Letter Drafter

Generates a demand letter based on case facts, medical records, and damages.

• Input: Case with medical records and bills
• Output: Draft demand letter with damage calculations

Medical Records Summarizer

Summarizes medical records into a structured timeline.

• Input: Medical records and treatment history
• Output: Chronological summary with diagnoses, treatments, and costs

Case Gap Analyzer

Identifies missing information or documentation in a case.

• Input: Current case data
• Output: List of gaps with recommended actions

How to Use

1. Open AI & Automate → Agents.
2. Find the agent for your task and click Run.
3. Select a case if the task needs its records. If the case does not exist yet,

choose Create a case and complete the form inside the dialog. The new case is selected when you return, with your additional input preserved.

4. Add any task-specific context in Additional Input.
5. Review the displayed credit cost and balance, then click Run Agent.
6. Follow the progress and review the result. Check Run History for the

recorded status and output.

See Creating a Case for the three-step form.

AI Credits

The run dialog shows the selected agent's credit cost and your remaining balance before you start. Costs vary by agent. Your firm's permissions, AI budget and configured providers also determine whether a run can start.

If a Run Does Not Start

Read the error shown in the dialog. If it concerns access, credits, budget or provider configuration, ask your firm administrator to resolve that condition. For another error, record its exact wording, the agent name and the approximate time for support. Check Run History before retrying so you can see whether a run was already created.

Output Quality

Review the output against the underlying case records and any source references provided. The information shown varies by agent; do not assume every result has citations or a confidence score. Review drafts and analysis before using them in case work or legal proceedings.

Run History

Open AI & Automate → Agents → Run History to inspect recorded runs, their status and available outputs. Start a new run when you need analysis of updated case data.