Prepared for Lengea — April 2026

Three MCP servers,
not one generic chatbot

Your four practice areas need different tools — medical spa clause libraries look nothing like employment agreement templates. Here's how I'd build a Claude-native legal practice infrastructure that scales with your firm.

Who I am

Alexander Girardet

AI Engineer — Claude MCP Specialist

I build production MCP servers and Claude plugin architectures. My work focuses on turning domain-specific workflows into Claude-native tools — the kind your team uses through conversation, not another dashboard to learn. I've built orchestration systems that coordinate specialist agents across CRM, data extraction, and content generation.

I've built this before

ICP Discovery Pipeline — a multi-agent system that coordinates prospect research, scoring, and CRM integration. Same architectural pattern: MCP servers connecting Claude to external APIs, structured data flows, and domain-specific tooling.

Previous build

ICP Discovery System

Orchestrator agent dispatches researcher agents that query external APIs (Exa, Google Places, Firecrawl), score prospects against configurable criteria, and commit results to a CRM (Attio) through a real-time dashboard. The system processes hundreds of prospects autonomously with full audit trails.

MCP Servers Claude Agents Attio CRM API Exa Search API Google Places API Firecrawl API Real-time Dashboard

Same pattern applies to Lengea: MCP servers connecting Claude to Clio, clause libraries, and legal research APIs — domain-specific tools your attorneys use through conversation.

Previous build

CRM & Ticketing Agent

Autonomous agent that centralizes CRM management and ticketing workflows — handles operational tasks across systems without manual intervention. Proof that AI-driven CRM automation works at the workflow level, not just the API level.

CRM Automation Autonomous Agent Ticketing Integration Workflow Orchestration Operational Management

Directly relevant to Lengea's client intake and matter management needs — the same CRM automation pattern applied to legal practice management.

Two ways to automate legal workflows

Most firms bolt AI onto existing tools. We build it into Claude where your team already works.

Standard approach

  • Zapier/Make — glue between apps
  • HotDocs — document assembly
  • Generic chatbot — basic Q&A
  • Separate dashboards — per tool

Multiple tools that don't talk to each other. Your team switches contexts constantly. Each tool has its own login, its own logic, its own limitations. AI is an add-on, not the interface.

Claude-native MCP servers

  • Clio MCP — CRM through Claude
  • DocGen MCP — clause-aware drafting
  • Research MCP — cited legal memos
  • One interface — Claude Cowork

Three purpose-built servers that give Claude deep access to your practice management, document templates, and legal research. Your attorneys interact with one interface — Claude — and it handles the rest.

The math

At 2,000+ clients across four practice areas, even small efficiency gains compound fast.

4,200+
Hours saved per year
16x
Year-one ROI
2 wks
First deliverable

Three MCP servers, phased delivery

Each server is independent but they compose together inside Claude Cowork Enterprise. Build one, prove value, expand.

Phase 1 — Weeks 1-4

Clio MCP Server

Connects Claude directly to your practice management system. Client intake, matter creation, conflict checks, task assignment, document uploads — all through conversation.

Tools

search_contacts create_matter run_conflict_check create_task upload_document Clio REST API OAuth 2.0

Phase 2 — Weeks 5-10

Document Generation MCP Server

Template engine with jurisdiction-aware clause libraries. Generate MSO agreements, compliance audit checklists, employment contracts — each customized by state and practice area.

Tools

generate_document get_clause suggest_clauses review_document Clause Library (Git) docx.js

Template coverage

MSO/MSA Agreements Medical Director Contracts Entity Formations HIPAA Compliance Audits Non-Compete Agreements Scope of Practice Memos

Phase 3 — Weeks 11-16

Legal Research MCP Server

AI-powered research tools — jurisdiction comparison, case law search, regulatory summarization. Generate cited legal memos in minutes instead of hours.

Tools

research_topic compare_jurisdictions generate_memo check_updates CourtListener API State Regulation DBs
All three servers compose — DocGen pulls client data from Clio, Research attaches memos to matters

How it works: new client intake

What currently takes 45-60 minutes becomes a 10-minute Claude conversation.

01

Attorney describes the new client

"New client — Dr. Sarah Park wants to open a medical spa in Texas. She needs entity formation and compliance review."

02

Claude runs conflict check via Clio MCP

Searches existing contacts and matters for conflicts. Returns clear or flags issues for review.

03

Creates contact and matter in Clio

New contact record, new matter tagged with practice area (Medical Spa Law), linked documents and intake checklist.

04

Assigns tasks to team

Paralegal gets MSO/MSA drafting task. Associate gets Texas scope-of-practice research. Calendar entries created for follow-ups.

05

Generates initial documents

DocGen MCP assembles a Texas MSO agreement draft from templates, selecting jurisdiction-specific clauses automatically. Uploaded to the matter in Clio.

What you get

Three MCP Servers

Clio integration, document assembly with state-specific clause libraries, and legal research — each installed in your Claude Cowork Enterprise environment, ready for your team to use.

AI-Readable Docs

Full documentation as markdown that Claude itself can read. Your team can ask Claude how any tool works, what it does, and how to extend it — no separate wiki to maintain.

All Source Code

You own everything. No vendor lock-in, no licensing fees, no dependency on me for changes. Your team can customize, extend, and maintain the servers directly.

Timeline

Phased delivery — prove value with Phase 1 before expanding.

W2

Clio API connected, first tools live

Contact search, matter creation, conflict checks working in Claude. Team can start testing.

W4

Phase 1 complete

Full Clio MCP server deployed with 10+ tools, documented, and tested with your team.

W10

Phase 2 complete

Document generation server with clause libraries covering your top practice areas and jurisdictions.

W16

Phase 3 complete

Legal research server with jurisdiction comparison, memo generation, and regulatory monitoring.

Want to get started?

Reply on Upwork and I'll send a detailed scope for the Clio MCP server within 24 hours.

I'd love to discuss which practice area to prioritize and how your team is currently using Claude.