// Intake systems for real estate + insurance

Answer every lead in minutes, not mornings.

AI-powered intake and follow-up that responds while your competitors sleep. Built for real estate teams and insurance agencies. Assessment first. Build second.

intake-agent.ts · live
// new lead · zillow · 11:42 pm
> parse  name · intent · budget ✓
> score  8.4 / 10 · hot
> route  sarah · listings team
> reply  sms + email · 0.8s
> crm    updated ✓
// waiting for next lead

Example intake agent run: a new lead arrives at 11:42 pm, is parsed, scored 8.4 out of 10, routed to the listings team, and answered by SMS and email in under one second with the CRM updated.

  • 18+years enterprise delivery
  • 2,500+hours building AI
  • F500AI POC in flight
  • 100%client-owned systems
  • Hilton Worldwide
  • American Airlines
  • MuleSoft Agent Fabric POC
  • Gateway Church build

Most AI projects fail before they start.

Wrong starting point

Businesses buy tools before they know what problem they're solving.

Generic builds

Templates and Zapier recipes that break the moment your workflow changes.

No one to own it

The freelancer disappears. The automation breaks. Someone manual-processes it again.

We fix all three. Starting with a direct audit of where you actually are.

What We Build

Everything starts with an Audit. Every Audit produces a roadmap. Every roadmap item is a scoped service.

Entry Point: Start Here

AI Opportunity Audit

From $500 · 5 days

Most businesses exploring AI don't have an AI problem. They have a clarity problem: too many options, no clear starting point, and a graveyard of tools that didn't stick.

Lead Intake & Follow-Up Automation

Learn More

Your lead pipeline has a response-time problem.

$750–$3,000

We build to your actual stack, whether that's GoHighLevel, HubSpot, Airtable, or a combination. No vendor lock-in to our preferred tools. Every build includes AI qualification and routing logic, automated follow-up, CRM logging, and a Loom walkthrough with written usage documentation.

Internal Workflow Automation

Learn More

Most businesses have ten or more manual workflows eating hours they can't get back.

$1,200–$4,000

We map your highest-cost manual workflows and build automations that eliminate the repetitive steps, keeping humans in the loop where judgment matters. Stack-agnostic. Built to integrate with what you already use.

AI Agent Build

Learn More

An AI that works inside your business: not a chatbot, a functional system.

$2,500–$6,000

We build purpose-specific AI agents, not general-purpose chatbots. Every agent is scoped to a defined job: lead qualification, internal Q&A, document processing, or customer support. Scope before build, every time.

AI Readiness Stack Setup

Learn More

AI only works if your data, tools, and workflows are ready for it.

$2,000–$5,000

Most businesses aren't AI-ready, and most vendors won't tell you that until after you've paid. This engagement gets your stack ready: clean data pipelines, connected tools, the right cloud infrastructure, and a documented architecture your AI investments can actually run on.

Recurring

AI System Maintenance & Growth Retainer

Automation breaks when APIs change, when your workflow evolves, or when a trigger stops firing. A retainer keeps everything running and adds capacity as your needs grow.

$500–$1,500/month

How We Work

1

Audit

We assess your workflows honestly. If automation isn't the right move yet, we'll tell you.

2

Roadmap

You get a prioritized action plan with specific tools, build estimates, and sequence.

3

Build

We build to your stack, your budget, and your actual workflow. Documented for your team.

4

Own It

You own everything we build. No vendor lock-in to us. No maintenance dependency.

We run our own business on the same systems we build for you. That's not a tagline: ask us to show you.

Talk to our AI

Built by someone who's done this at every scale.

Paul Dunphy has spent 18+ years running complex technical programs for organizations including Hilton Worldwide and American Airlines, and is currently leading an enterprise agentic AI proof-of-concept for a Fortune 500 client using MuleSoft Agent Fabric.

In parallel, he's spent 2,500+ hours since 2022 building AI systems from scratch: RAG pipelines, self-hosted automation infrastructure, and AI-powered systems for small businesses and nonprofits.

Today that work is concentrated on two sectors where the gap between response speed and closed business is largest: real estate teams where a 10-minute delay means a lost client, and insurance agencies where slow follow-up is the primary reason prospects go elsewhere.

Red Phoenix AI exists because most AI consultants either advise without building, or build without thinking. We do both, and we show our work.

Built on the systems we sell

We don't advise on AI. We run it.

Every recommendation comes from production experience. Here's what exists today.

Pro bono

Gateway Church: AI volunteer matching system

Designed and led the build of an AI-powered volunteer matching system for a large faith-based nonprofit. The system surfaces ranked recommendations; people make the final call. Responsible AI from the start: no black-box automation where human judgment matters.

Delivered 2026

Infrastructure

Self-hosted n8n on DigitalOcean

We deploy and maintain our own n8n automation infrastructure on DigitalOcean. Full ownership, no per-task vendor fees, no platform lock-in. The same setup we recommend to clients where data privacy or cost control matters.

2022 – present · live

Architecture

Personal RAG system

Built a custom retrieval-augmented generation system from scratch: OpenAI for the language layer, Pinecone for vector storage, Obsidian as the knowledge base, n8n for orchestration. The same architecture scales to SMB internal AI assistants and document Q&A tools.

Architecture diagram only · no client data

Enterprise

Fortune 500 MuleSoft Agent Fabric POC

Currently leading an enterprise agentic AI proof-of-concept using MuleSoft Agent Fabric for a Fortune 500 client. Scope includes stakeholder alignment, risk management, cross-functional coordination, and structured delivery across distributed teams.

Active · NDA · reference only, no specifics

Not sure where to start? Talk to our AI.

Answer a few questions and we'll tell you honestly whether AI automation makes sense for your business right now, and what the right starting point is.

PhoenixAI Intake Assistant
Hey, I'm Phoenix, Red Phoenix AI's intake assistant. I can help figure out whether AI automation makes sense for your business, and where to start if it does. To give you an honest answer, I need to ask a few quick questions. Sound good?

Powered by Claude

Prefer to talk directly? Book a 20-minute call

Get in touch

Prefer email over a chat bot? Send a message and we'll reply within one business day.