n8n Workflow Automation for Legal Leads

Summary WHAT TRYING TO DO OPEN TO FEEDBACK: The goal Get you retained as an expert witness more often by finding relevant lawsuits the day they're filed and letting the attorney know you're available — before they've gone looking for an expert. What we're building A small automated pipeline that runs every weekday morning: Trellis (a court-records service) emails you alerts about newly filed Texas cases matching your searches (med-mal, epidural injections, RFA, stimulators, radiculopathy, CRPS, concussion, trucking/spine injuries). n8n (a no-code automation tool) reads those alert emails, pulls out each case, and skips any you've already seen. Claude (AI) reads the case details and scores it 0–100 for fit with your practice, identifies the plaintiff attorney and firm, and writes a two-sentence summary. It never gives a medical opinion. Hunter finds and verifies the attorney's email address. Claude writes a short, professional email using your fixed template ("I became aware of this matter… I'm board-certified in FM, Sports, and Pain… happy to speak if you need an independent expert"). The email is saved as an Outlook draft and logged in Airtable (a simple spreadsheet-style CRM). Nothing sends on its own. You read the draft on your phone and click Send or delete it. A second workflow watches your Sent folder and inbox, marks each case Sent/Replied/Retained in Airtable, and notifies you when an attorney responds. It also blocks re-contacting the same attorney within 90 days and honors unsubscribes. DESCRIPTION I'm a physician setting up a small lead-monitoring automation. The workflows are ALREADY BUILT and exported as n8n JSON files (23 nodes and 10 nodes), with a written setup guide. I need an experienced n8n freelancer to import them, wire up the credentials, build the matching Airtable table, run a successful test, and hand it back working. No new logic needs to be designed. What the workflows do (high level): • Workflow A: reads alert emails from a court-monitoring service in my Outlook inbox → extracts case data with the Claude API (HTTP Request node) → dedupes against Airtable → scores the case → looks up a contact email via Hunter.io → drafts an email → saves the record to Airtable → creates an Outlook DRAFT (nothing is auto-sent; I approve every email manually). • Workflow B: hourly; watches my Sent folder and inbox and updates Airtable statuses, then emails me when a reply arrives. DELIVERABLES 1. Both workflows imported into my n8n Cloud account and all credentials attached (Microsoft Outlook OAuth2, Airtable token, Anthropic header auth). 2. An Airtable base built to the exact field spec in my setup guide (about 25 fields, one table, one view). 3. All placeholders in the JSON replaced (Airtable base ID, Hunter API key, my email, my contact details in one prompt). 4. One successful end-to-end test of Workflow A using a real alert email, with a draft appearing in my Outlook Drafts folder and a row in Airtable, plus a clean test of Workflow B. 5. A 15-minute screen-share walkthrough (Zoom/Teams) showing me how to change the score threshold and how to read errors, plus a short list of any changes you made and why. WHAT I PROVIDE • The two n8n JSON files and a step-by-step setup guide • Accounts and API keys for n8n Cloud, Airtable, Anthropic, Hunter, and Microsoft 365 (I'll enter keys myself on a screen-share, or provide temporary keys I rotate afterward) • Prompt availability to answer questions REQUIREMENTS • Demonstrated n8n experience (link to portfolio or describe 2 comparable builds) • Comfortable with n8n HTTP Request node, JSON expressions, Code nodes (JavaScript), the Microsoft Outlook node, and the Airtable node v2 • Familiar with the Anthropic Messages API (or any LLM API — it's a simple POST) • Good written English; will not modify the email-drafting prompt's content without asking • Confidentiality: the data involves attorney contact information and my business development; you agree not to retain or reuse it BUDGET & TIMELINE Fixed price, $250 (open to proposals in the $150–$400 range based on experience). I expect 2–4 hours of work. Ideally completed within 3 days of hire. Small bonus if it works on the first test and the handoff is clean. SCREENING QUESTIONS (please answer in your proposal) 1. Which n8n version/hosting have you used most recently (Cloud or self-hosted)? 2. Have you connected the Microsoft Outlook OAuth2 node before? Any gotchas? 3. Briefly describe one workflow you built that called an LLM API through the HTTP Request node and parsed JSON from the response. 4. What's your availability this week? NOTE Please do not send generic proposals. Start your proposal with the word "Draft" so I know you read this.