Nettlio
Confidential · The Hang Loose Hut
The Hang Loose Hut · AI Integration Proposal · Prepared for Sean Zent

Every inbound order, handled by one person and AI.

The Hang Loose Hut runs on inbound. Today three people answer every sizing, pricing, and timeline question by hand over text. We put Claude on the front line, trained on your real conversations, so one sales-minded person carries the volume and every lead gets a faster, on-brand reply.

Three reps texting inbound leads by hand
→
One reviewer plus Claude drafting the bulk of replies
iMessage threads living in one person's phone
→
Sales Message and HubSpot, with Claude trained on your FAQs
Pricing, policies, and FAQs scattered across Sheets, PDFs, and Docs
→
One structured data layer the AI can actually use
What we heard
We have three people doing the job I feel can be done with one person.
Sean Zent · The Hang Loose Hut
Inbound leads are answered by hand over text today. With the right setup, one sales-minded person plus AI can carry the same volume and reply faster, while keeping the human touch your niche expects.
We're moving off iMessage to Sales Message and HubSpot. I want to set it up right so we don't backtrack and scrap it.
The Hang Loose Hut · The migration
This move is the moment to build the foundation properly. A unified data layer and a human-in-the-loop workflow mean the AI compounds over time instead of getting thrown away.
I uploaded everything and it's weird it didn't just know. But it's like a human, it still needs training.
The Hang Loose Hut · Training the AI
Document uploads alone do not make a good responder. Claude learns from examples of your actual conversations. That is where accuracy and your short, priced, on-brand tone come from.
How Nettlio works

Get under the hood. Don't bolt AI on top.

We don't drop a chatbot on your inbox and hope. We build a structured data layer from your real business knowledge first, then put AI on top of it with a human in the loop. You start with one section, inbound leads, prove it works, then expand. It is the same approach we use to make companies AI-native, from small businesses to family offices.

Data layer
The foundation. Your pricing tool, FAQs, order minimums, and policies pulled out of scattered Sheets and PDFs into one structured source the AI can reason over.
Human in the loop
One sales-minded person reviews and approves replies at the start. The AI catches their errors, they catch the AI's. Autonomy grows only as accuracy earns it.
Trained on your conversations
Real example threads, not just documents. That is what makes replies sound like your team: short, quick, and priced, not long AI essays.
Built to expand
An agentic layer, not one-off automation. Once inbound leads work, the same foundation extends to factory comms and beyond.
The rollout

Start with inbound. Prove it. Then expand.

Phase 1 is a small, contained project with a clear outcome. Phase 2 and the wider rollout follow only once Phase 1 is working, so nothing gets built ahead of the results that justify it.

1
Phase 1 · Signing now · live in weeks
Inbound Lead Co-Pilot
Claude drafts replies to inbound texts inside Sales Message, trained on your FAQs and real conversations. One person reviews and sends.
Details
What ships
01
A structured data layer from your existing knowledge
Sizing, pricing, order minimums, production timelines, and refund policy pulled out of the Google Sheets, PDFs, and Docs into one structured source Claude can actually use.
02
Claude trained on your real conversations
We load example inbound threads, the typical questions and how your team answers, so replies sound like The Hang Loose Hut: short, quick, and priced, not long AI-style essays.
03
A drafting workflow inside Sales Message
Claude drafts the reply, your reviewer approves, edits, or sends. Simple FAQ answers move fast. Nuanced or high-value threads get escalated to the person.
04
HubSpot connected
Every lead and conversation lands in HubSpot, so context is unified and nothing lives in one rep's phone.
05
Oversight setup and a handoff session
A simple review setup for the person managing Claude, plus a training session so your team owns it and can keep feeding it examples.
What we need: access to your Sheets, PDFs, and Docs (pricing, FAQs, minimums, policies), Sales Message and HubSpot access, and a batch of past inbound threads to train on.
2
Phase 2 · Scoped after Phase 1
Factory and Production Comms
Automate follow-ups with the factory off your Google Sheet production tracker, once inbound leads are running cleanly.
Details
What it covers
01
Production status follow-ups
Read from the Google Sheet that tracks orders, then chase the factory manager automatically: confirm a sample, confirm an order ships tomorrow, flag anything behind schedule.
02
Time back for management
The repetitive checking-in on production moves to AI, so the team only steps in on the exceptions that need a human.
When: after Phase 1 proves out. We scope and price it together once inbound leads are mastered.
+
Horizon · As the partnership matures
Toward AI-Native Operations
The same data-layer foundation extends to other parts of the business as you see the impact.
Details
Where it can go
→
Other departments, one at a time
Accounting, marketing, sales, and legal processes are natural next steps. Each one becomes more efficient on top of the structured foundation, rather than a separate bolt-on tool.
→
Support as the model and business evolve
AI is moving fast. Ongoing support keeps your setup current as the models improve and as The Hang Loose Hut grows.
Roadmap

Phase 1 in the first weeks. Everything after compounds on it.

A simple, honest sequence. The inbound co-pilot is the only committed scope. Phase 2 and the horizon are placed where they would land, and scoped only when we get there.

Proposed sequence · confirmed at kickoff
Track
M1
M2
M3
M4
M5
M6
Phase 1 · Inbound Co-Pilotthe committed scope
Data layer + FAQ setup
Train + draft live
Live, human-in-the-loop, tuned weekly
Human oversightone reviewer
Reviews and approves every reply. Autonomy grows as accuracy earns it.
Phase 2 · Factory commsHoldings of production
Scoped after Phase 1 proves out
Horizonother departments
Accounting, marketing, sales, legal

Timeline is indicative. Sean noted Phase 1 is a small project to start. We confirm the exact sequence at kickoff.

Pricing · Phase 1

A clean, fixed scope to start.

One line on this proposal: the Inbound Lead Co-Pilot, at a fixed one-time price. Phase 2 and ongoing support are a separate conversation we have once Phase 1 is working.

A small, contained first project. No retainer required to start. We earn the next phase by proving this one.

Separate · scoped after Phase 1
Phase 2 and ongoing support

Once inbound leads are running cleanly, Phase 2 automates factory and production follow-ups off your Google Sheet tracker: confirming ship dates, chasing delayed samples, flagging orders behind schedule. Ongoing support is there as the models improve and your business evolves.

Scoped togetherafter Phase 1

No long contract to start. We price Phase 2 and any ongoing support against the results Phase 1 delivers.

The shape of the engagement
Phase 1
$3.8k one-time
Phase 2 + support
TBD
Phase 2 and ongoing support are scoped and priced once Phase 1 is live and proven.
Next steps

You review. We build.

Phase one is a small, contained project with a clear outcome: one person plus Claude handling your inbound. Prove it here, then we look at the rest of the business.

Questions or a quick call: jens@nettlio.com · nettlio.com

1
Sean reviews this plan Take a look and share your thoughts. Happy to adjust the scope on a quick call.
2
Confirm Phase 1 Greenlight the $3,800 fixed scope and we book a kickoff.
3
Share your knowledge The Sheets, PDFs, and Docs with pricing, FAQs, minimums, and policies. Plus a batch of past inbound threads to train on.
4
Sales Message and HubSpot access So Claude can draft inside your inbox and every lead lands in the CRM.
5
Build the data layer and train We structure your knowledge and train Claude on your real conversations.
6
First drafts go live Your reviewer starts approving AI-drafted replies, and we tune weekly from real threads.