Legal Services

Legal Document Service (Process Serving)

Legal Services / Document Service (Process Serving)

We turned a scan of a handwritten proof of service into a court-format server's affidavit, ready to print and sign — within minutes, without a single clerk typing a word.

Google Gemini Vision API
n8n
Supabase
PDF Generation Library
Email/IMAP
Local folder + scanner with a feeder
Challenge

The Challenge

According to the client's discovery account, hundreds of proofs of service returned from the field each day and four to five process servers waited for affidavits to be prepared. Most forms included four handwritten fields — date, time, recipient and signature — that staff manually entered into the required document format. This volume is client-reported, not a measurement made by us. The verifiable project requirement was Hebrew handwriting OCR with human review and PDF output, without promising perfect accuracy or bypassing the handwritten-signature step.

Solution

The Solution

We built an end-to-end automated processing line. The scanned proof of service comes in through a simple channel (a web app, email or a local folder), and Google Gemini Vision automatically extracts both the printed text and the Hebrew handwriting — the server's name, the recipient, who received the service, date and time. The data goes through a quick human approval check (human-in-the-loop) of a few dozen seconds to ensure accuracy in a legal document, and is then automatically inserted into a server's-affidavit template in the exact court format and exported as a high-quality PDF, ready to print and sign by hand. A connection to a scanner with a feeder and batch-processing logic lets you throw in 50 pages at once and get back a stack of affidavits — with automatic folder monitoring and error alerts. Every affidavit is saved in an archive searchable by date and server. The flow is orchestrated in n8n, with storage on Supabase.

Our Approach

An intake channel for scanned proofs of service — a web app, email and a local folder
Google Gemini Vision integration for OCR of Hebrew handwriting and printed text
Fixed field extraction by the proof-of-service rubrics (server, recipient, receiver, date/time)
A data review and approval interface (human-in-the-loop) before the document is created
Automatic creation of a server's affidavit in an exact court template and a PDF export ready to print and sign
An affidavit archive searchable by date/server
A connection to a scanner with a feeder and batch processing (50+ pages at once), including folder monitoring and error alerts

Technologies Used

Google Gemini Vision APIn8nSupabasePDF Generation LibraryEmail/IMAPLocal folder + scanner with a feeder

The service behind this project

Business Automation
Results

The Results

The project was delivered and included Hebrew-handwriting OCR, a human-review queue, batch processing and PDF output in the defined format. The starting volume is based on the client's account. Post-launch handling time, error rate and hours saved were not measured in the project data, so we do not claim that processing finished within minutes or that errors fell by a quantified amount.
Schedule a Meeting

Ready to achieve similar results?

Let's meet and explore how your business can achieve meaningful transformation

View More Case Studies

Schedule a Meeting

Let's meet and explore how your business can achieve meaningful transformation