All templates By NappAI team Legal Contract Analyzer
Automation, analysis, and summary of complex legal documents, acting as an AI assistant for lawyers, managers, and anyone who needs to quickly understand the key points of a contract.
graph TD
%%{init: {'theme': 'mc','layout': 'elk'}}%%
GDriveFilesComponent-om3qs[<div><img src="/_astro/google_drive.wKmDsV2c.svg" style="height: 20px !important;width: 20px !important"/></div>Google Drive]
style GDriveFilesComponent-om3qs stroke:#a170ff
ParseData-bpxzj[<div><img src="/_astro/braces.Djq0PW4_.svg" style="height: 20px !important;width: 20px !important"/></div>Parse Data]
style ParseData-bpxzj stroke:#a170ff
LanggraphReactAgent-r7y0s[Asistente Legal]
style LanggraphReactAgent-r7y0s stroke:#a170ff
ChatOutput-fu8q8[<div><img src="/_astro/messages-square.BaSDmT6g.svg" style="height: 20px !important;width: 20px !important"/></div>Chat Output]
style ChatOutput-fu8q8 stroke:#a170ff
OpenAIModel-mog55[<div><img src="/_astro/openAI.BhmuxEs3.svg" style="height: 20px !important;width: 20px !important"/></div>OpenAI]
style OpenAIModel-mog55 stroke:#a170ff
GDriveFilesComponent-om3qs -.- ParseData-bpxzj
linkStyle 0 stroke:#a170ff
ParseData-bpxzj -.- LanggraphReactAgent-r7y0s
linkStyle 1 stroke:#a170ff
OpenAIModel-mog55 -.- LanggraphReactAgent-r7y0s
linkStyle 2 stroke:#a170ff
LanggraphReactAgent-r7y0s -.- ChatOutput-fu8q8
linkStyle 3 stroke:#a170ff
Legal Contract Analyzer
The Legal Contract Analyzer automates the extraction, interpretation, and summarisation of legal documents. By integrating Google Drive for file ingestion, a text‑parsing routine, an advanced Langgraph agent powered by an OpenAI model, and a chat‑style output component, the workflow delivers concise, bullet‑point summaries that highlight key contractual elements. This streamlines legal review, enabling lawyers and managers to grasp essential points quickly and accurately.
⚙️ Main Features
- Automated File Retrieval – Pulls legal documents directly from Google Drive, supporting a variety of file types.
- Text Normalisation – Converts binary or structured document data into clean, plain‑text format ready for analysis.
- AI‑Driven Summarisation – Uses a Langgraph React Agent with an OpenAI model to read contracts and generate simple, bullet‑point summaries.
- Chat‑Friendly Presentation – Displays the AI‑generated summary in a conversational format, suitable for review or further discussion.
- Extensible Tooling – The agent can be extended with additional tools or memory management features for more complex workflows.
🔄 Workflow Steps
| Component Name | Role in the Workflow | Key Inputs | Key Outputs |
|---|---|---|---|
| Google Drive | Retrieves the selected legal document from a Google Drive folder or file ID. | File ID or Folder selection (via UI). | Raw file data (binary or structured). |
| Parse Data | Transforms raw file data into plain‑text suitable for natural‑language processing. | Raw file data from Google Drive. | Extracted text (Message). |
| OpenAI Model | Provides the language‑model backend that the agent uses to generate a summary. | Model name (gpt-4o-mini), API key, configuration settings (temperature, max tokens). |
Language‑model instance usable by the agent. |
| Legal Assistant (Langgraph React Agent) | Orchestrates the summarisation task: receives the document text, invokes the OpenAI model, and produces a concise summary. | Document text from Parse Data, Language‑model instance from OpenAI Model. | Summary message (Message). |
| Chat Output | Presents the summary to the user in a chat‑style interface. | Summary message from the Legal Assistant. | Displayed chat message in the playground. |
🧠 Notes
- The workflow assumes the presence of valid Google Drive API credentials and OpenAI API keys; missing credentials will halt execution.
- File size limits imposed by Google Drive and the OpenAI model (e.g., token limits) may constrain the length of contracts that can be processed in a single run.
- The OpenAI model is invoked in JSON mode by default; any required output format must be specified in the system prompt.
- The agent’s memory and summarisation features can be toggled via configuration flags, allowing for either stateless or stateful summarisation across multiple documents.
- The chat output component is purely visual; it does not alter the underlying data or trigger additional actions.
- If the input file contains complex formatting (e.g., tables, legal clauses with nested bullet points), the parsing step may need to be customised to preserve meaningful structure.