Ham Bone AI
Engagement model

Four phases.
Thirty days to live.

We do not sell discovery decks. The engagement starts with us on your floor and ends with your team running systems they can extend without us.

01
Week 1

Walk the floor

We come to you. We watch the work happen, talk to the people doing it, and write down every place a process exists only because it always has.

Phase output
  • On-site observation across shifts
  • Interviews with floor, office, and leadership
  • Systems inventory: ERP, sheets, drives, email
  • Written map of where work actually stalls
Phase 1
02
Week 2

Find the small pot

Every bottleneck gets sized: hours burned, dollars leaked, risk carried. Then we rank them by what AI can actually fix this quarter.

Phase output
  • Bottleneck register with hours and cost
  • Feasibility and data-readiness scoring
  • Ranked build queue with expected return
  • Explicit list of what we will not automate
Phase 2
03
Weeks 3–8

Build in your environment

We build against your real files, your real ERP, your real people — starting with one workflow that ships and gets used, not a pilot that gets demoed.

Phase output
  • First workflow live inside 30 days
  • Approval gates on anything that writes
  • Role-based access from day one
  • Weekly working sessions, no status theater
Phase 3
04
Ongoing

Hand over the pot

Your team learns to run and extend what we built. Templates, documentation, and training so the capability stays after we leave.

Phase output
  • Operator and admin training
  • Reusable workflow templates
  • Documented ownership and escalation
  • Optional quarterly optimization retainer
Phase 4
Operating principles

Rules we
don't bend.

P/01

Nothing writes without approval

AI prepares, humans commit. Every workflow that touches an ERP, a marketplace, or a shared file stages its change for review first.

P/02

One live system beats five demos

We ship the first workflow into real use before starting the second. Adoption is the metric, not deployment.

P/03

Your data stays yours

We build inside your accounts and your tenancy. No proprietary black box you cannot open after we leave.

P/04

We tell you when AI is wrong

Some bottlenecks are staffing, layout, or pricing problems. We will name them instead of selling around them.

What you keep

Deliverables, in plain terms.

  • 01Written bottleneck register with hours and dollars attached
  • 02Ranked build queue with expected return per workflow
  • 03Working AI systems inside your own accounts
  • 04Role and permission model for company-wide AI use
  • 05Approval gate design for anything that writes to live systems
  • 06Reusable workflow templates your team can clone
  • 07Operator and admin training sessions
  • 08Documented ownership, escalation, and maintenance plan

Phase one starts with a walk
around your building.

Schedule the floor walk