02 / Selected work

Proof, not promises.

Production systems running real operations for law firms and service businesses — each tied to a number, not a demo.

US Law Firm · Legal AI Automation

AI Email Automation for a US Law Firm

7–10 min1–2 min
per email≈80% faster
1
Problem

A US law firm handled high volumes of client, court, and counsel email by hand — reading each, pulling matter context from Clio, drafting, then polishing with AI. Routine replies took 7–10 minutes.

2
What I built

An n8n automation linking Outlook (Microsoft Graph) and Clio. An AI agent runs 24/7: filters irrelevant mail, pulls matter context automatically, and drafts a context-aware reply ready to review and send.

3
Result

Replies now take 1–2 minutes to review and send instead of 7–10 — about 80% faster — with irrelevant mail filtered out before it reaches the queue.

n8nMicrosoft GraphClioGPT-4oOutlook
Read the full case study
Results

≈80%

faster replies

1–2m

per email

24/7

always on

100%

matter context

Workflow
Incoming
AI agent
Draft
Sent
US Personal-Injury Firm · Legal Document Automation

AI Demand-Letter Generation for a Personal-Injury Firm

3–4 hrs~25 min
per demand≈85% faster
1
Problem

Personal-injury demand letters are built by hand — pulling facts from intake notes, medical records, and billing, then structuring liability and damages into a persuasive package. It's detail-heavy work that ties up paralegals and attorneys on every case.

2
What I built

An AI pipeline that ingests case documents, extracts the facts and damages that matter, and drafts a structured, attorney-ready demand letter in the firm's format — consistent every time, built for review rather than from scratch.

3
Result

Attorneys start from a complete first draft instead of a blank page — case facts assembled and formatted automatically, so the work shifts from assembling to reviewing.

GPT-4oClaudePythonDocument AIRAG
Read the full case study
Assembly pipeline
Intake notes
Medical records
Billing
AI agent
Demand
Demand letter · draft
RE: DEMAND FOR SETTLEMENT
TOTAL DAMAGES$—
attorney-ready first draft
Professional Services · CRM Automation

AI-Connected CRM & Workflow Automation

~8 hrs/wk<1 hr/wk
manual upkeep≈90% less
1
Problem

The business ran its pipeline across Stacker and Airtable, with updates, hand-offs, and follow-ups done manually. Data drifted out of sync between tools and routine steps quietly ate the team's day.

2
What I built

n8n automations wiring Stacker and Airtable together with an AI layer: records sync automatically, routine actions fire on triggers, and the AI handles the judgment-call steps — enrichment, routing, and drafting — in between.

3
Result

The CRM stays current on its own and the manual busywork between tools is gone — the team works the pipeline instead of maintaining it.

n8nAirtableStackerClaude / GPT-4oWebhooks
Read the full case study
Sync flow
Airtable
Stacker
n8n
AI
CRM
CRM · live
PIPELINEsynced just now
Lead · enriched
Stage · synced
Follow-up · queued

Losing hours to work an AI system could handle?

Tell me about the process. I'll tell you straight whether it's worth automating — no pitch, no obligation.

Book a call