DEADHEADDESK · FORM DD-00 · REV A THE DESK IS ON · --:--:-- LOCAL
Form DD-03 · The managed service

Not software you babysit.
Not an agency that vanishes.

Most "AI for brokers" comes in two flavors, and both fail the same way: nobody is watching after the demo. DeadheadDesk is the third option — a narrow service that is tested nightly, maintained forever, and gated by your reps on every send.

Option A · software you babysit

"It demos great. Then a carrier writes 'chi to atl lmk' at 1 AM."

You buy a seat-licensed AI tool. It prices clean emails fine — and quietly mangles the messy ones. In published industry surveys, 61% of teams running off-the-shelf AI tools report accuracy complaints. The vendor's answer is a knowledge-base article.

Option B · the agency build

"They built it, billed us, and moved on."

An automation shop wires up a workflow and disappears. Six months later a TMS API version bumps and quotes silently stop logging. Roughly 74% of companies report rolling back or stalling at least one AI initiative — most after the builder was gone.

FIGURES: PUBLISHED INDUSTRY RESEARCH ON ENTERPRISE AI ADOPTION, 2024–25. AGGREGATED AND DIRECTIONAL — NOT OUR DATA. THAT'S THE POINT: NOBODY WAS WATCHING.

The third option

A managed desk: two narrow agents, a nightly eval harness, a human approval gate on everything, and a named person on the hook when it breaks. The rest of this page is the contract, in plain English.

The first 30 days

One inbox. One lane. Four weeks.

The pilot is $2,500, credited in full against the $7,500 setup if you continue. Each week ends with a deliverable you can hold in your hand. Click through the weeks.

WEEK 1 · DISCOVERYDAYS 1–7 · NOTHING TOUCHES A CUSTOMER

One inbox, one lane, your margin rules on paper.

We pick a single quotes inbox and a single corridor with real volume. Then we sit with your best rep and write down how they actually price — floors, targets, fuel posture, customers who get the friendly number. If a rule lives only in someone's head, this is the week it gets written down.

Scope locked1 INBOX · 1 LANE
Margin-rule interviews2–3 SESSIONS
Historical email pull90 DAYS, SANITIZED
TMS accessREAD-ONLY TO START
Baseline measuredYOUR CURRENT REPLY TIME

Deliverable · end of week 1

FORM DD-11 — MARGIN RULE MAP ──────────────────────────── LANE: CHI → ATL · DRY VAN FLOOR ............. 12% MIN MARGIN TARGET ............ 15% FUEL .............. ALL-IN QUOTES KEY ACCOUNTS ...... 2 FLAGGED (REP-ONLY PRICING) ESCALATE ALWAYS ... MULTI-STOP · NEW SHIPPER · HAZMAT · ANYTHING ODD + BASELINE SHEET: YOUR MEDIAN REPLY TIME, MEASURED — THE NUMBER WEEK 4 IS JUDGED AGAINST.

CONTENTS ILLUSTRATIVE — YOUR MAP USES YOUR RULES.

WEEK 2 · INTEGRATIONDAYS 8–14 · STILL NOTHING CUSTOMER-FACING

Email webhooks, TMS connector, rate API.

The plumbing week. Inbound email starts flowing to the agent through webhooks. The TMS connector (Aljex, Tai, McLeod, Turvo, or Ascend) gets wired and tested against sandbox loads. DAT or Greenscreens rate pulls come online. Your historical emails are seeded into the eval corpus — the test set everything is graded against, forever.

Email webhooksLIVE, LISTEN-ONLY
TMS connectorALJEX / TAI / McLEOD / TURVO / ASCEND
Rate feedDAT · GREENSCREENS
Eval corpus seededFROM YOUR REAL EMAILS
Customer contactNONE. ZERO.

Deliverable · end of week 2

FORM DD-12 — INTEGRATION SIGN-OFF ──────────────────────────── [x] INBOUND EMAIL → AGENT (WEBHOOK) [x] TMS LANE-HISTORY READS VERIFIED [x] RATE API: LIVE PULLS MATCH PORTAL [x] TMS WRITE PATH TESTED (SANDBOX ONLY) [x] EVAL CORPUS: SEEDED + LABELED [x] CREDENTIALS: SCOPED, REVOCABLE, LOGGED SIGNED: YOUR OPS LEAD + J. MORGAN NO SIGN-OFF, NO WEEK 3.

CHECKLIST FORMAT — ACTUAL ITEMS VARY BY TMS.

WEEK 3 · SHADOW MODEDAYS 15–21 · DRAFTS ONLY, SILENTLY

The agent drafts on your real email. Nobody outside sees a word.

Every real quote request that hits the inbox gets a silent draft. Your reps keep working exactly as before — and grade the drafts: would have sent, would have edited, dead wrong. Every correction goes back into tuning. The nightly regression suite starts running against the corpus this week and never stops.

Drafts producedON REAL INBOUND, SILENT
Rep grading~2 MIN/DAY PER REP
Tuning loopCORRECTIONS → RULES → RE-TEST
Nightly regressionSTARTS · RUNS FOREVER

Deliverable · end of week 3

FORM DD-13 — SHADOW SCORECARD ──────────────────────────── DRAFTS PRODUCED ........... 117 GRADED "WOULD SEND" ....... 79% GRADED "MINOR EDIT" ....... 16% GRADED "WRONG" ............ 5% → ALL TRACED + FIXED, RE-TESTED ESCALATIONS (CORRECT) ..... 14/15 MARGIN-RULE VIOLATIONS .... 0 (HARD FAIL IF EVER ≠ 0)

SAMPLE DATA ILLUSTRATIVE SCORECARD — YOURS SHOWS YOUR GRADES.

WEEK 4 · LIVE, GATEDDAYS 22–30 · REPS APPROVE EVERY SEND

Live — with a human gate on every single send.

Drafts now land in an approval queue. A rep clicks approve, edit, or reject; nothing reaches a customer without that click. At day 30 we sit down with the before/after numbers against the Week 1 baseline and make a go/no-go call together. If it's no-go, you walk — the pilot ends at $2,500 and no setup fee is billed.

SendingREP-APPROVED, EVERY TIME
TMS loggingEVERY QUOTE, AUTOMATIC
Go / no-goMETRICS VS. BASELINE, NOT FEELINGS
If go$2,500 CREDITS TO $7,500 SETUP

Deliverable · day 30

FORM DD-14 — GO / NO-GO SHEET ──────────────────────────── MEDIAN REPLY: BASELINE VS. NOW AFTER-HOURS REQUESTS ANSWERED DRAFT APPROVAL-WITHOUT-EDIT RATE ESCALATION PRECISION QUOTES LOGGED TO TMS (SHOULD BE ALL) REP TIME SAVED (SELF-REPORTED) DECISION: GO / NO-GO — YOURS. NO-GO = WE PART FRIENDS. YOUR DATA AND CORPUS LABELS GO WITH YOU.

THE TEMPLATE IS THE PROMISE: HARD NUMBERS DECIDE.


The monitoring layer

Every night at 2 AM, the desk takes its own physical.

This is the part nobody else sells, and the part that matters most. A regression suite built from real freight emails runs nightly. Drift alarms watch confidence trends. If extraction quality slips, we get paged — you get a line item in the weekly report, not a surprise in your P&L.

NIGHTLY REGRESSION · 02:00SAMPLE DATA
SuiteCasesPassStatus
Quote extraction41299.0%PASS
Margin-rule compliance268100%PASS · HARD GATE
Carrier reply parsing39097.7%PASS
Escalation triggers14599.3%PASS
Double-broker red flags61100%PASS · HARD GATE
New: TriHaul dispatch format1283.3%⚑ PATCHING

MARGIN-RULE AND FRAUD SUITES ARE HARD GATES: ANYTHING UNDER 100% BLOCKS THE AGENT ON AFFECTED PATTERNS UNTIL FIXED.

DRIFT & ALERT LOGSAMPLE LOG
02:00:14 NIGHTLY SUITE START · CORPUS v38 (1,276 LABELED EMAILS) 02:07:41 PASS 1,262/1,276 (98.9%) BASELINE 98.7% · WITHIN BAND 02:07:42 DRIFT CHECK · EXTRACTION CONF 7-DAY MEAN 0.94 → 0.93 · OK 03:11:09 ⚑ ALERT · CARRIER "TRIHAUL" REPLY FORMAT CHANGED PARSE CONF DROPPED TO 0.61 03:11:10 AFFECTED PATTERN ROUTED TO HUMAN QUEUE · NO SILENT GUESSING 03:11:11 PAGE → ON-CALL (J. MORGAN) TICKET DD-218 OPENED 09:32:55 PATCH DEPLOYED · RE-RUN 12/12 PASS · NOTED IN WEEKLY REPORT
EXTRACTION CONFIDENCE · 6-WEEK TRENDSAMPLE DATA
WEEK -5
0.88
WEEK -4
0.90
WEEK -3
0.91
WEEK -2
0.93
WEEK -1
0.92 ⚑ DIP TRACED: TRIHAUL
THIS WEEK
0.94

MEAN EXTRACTION CONFIDENCE ON LIVE TRAFFIC. A DIP IS A TICKET, NOT A MYSTERY. ALL VALUES SAMPLE DATA.

Human in the loop

Autonomy is earned, never enabled.

Day one, your reps approve everything. Months later — if the numbers hold and you say so — routine lanes can earn auto-send. You control the ladder, rung by rung, and you can step back down any time.

THE TRUST LADDERCLIENT-CONTROLLED, REVOCABLE
RungWhat the agent doesWhen
0Approve everything. Every draft held for a rep's click. This is day one and stays default.DAY 1
1One-click flow. Same gate, faster queue. The agent learns from every edit your reps make.WEEKS IN
2Earned auto-send on specific routine lanes you designate — only after sustained accuracy on those lanes, only with your signature, revocable instantly.MONTHS IN
×Full autonomy across the desk. Not a rung. We don't sell it.NEVER
WHAT IS NEVER AUTONOMOUSNO EXCEPTIONS
ActionWhy
Sending quotes on day oneTrust is measured first, granted second.
Pricing for a new shipperNo history, no pattern, no auto-send. A rep owns the first impression.
Multi-stop pricingStop-off fees and liability need human judgment. The agent escalates with full context instead.
Carrier onboardingVetting a carrier is a fraud surface. Humans only.
Anything fraud-adjacentDouble-brokering signals, identity mismatches, changed remit-to details — these page a person, always.

The client artifact

The weekly report, denominated in dollars.

Every Monday you get this document. Not dashboards, not vibes — what the desk did last week, in loads and dollars. Toggle between two sample weeks to see how it reads as trust is earned.

SAMPLE REPORT — ALL NUMBERS ILLUSTRATIVE
FORM DD-07 · WEEKLY REPORT · PILOT WEEK 4SAMPLE REPORT

Summary · week of (sample)

MetricThis weekBaseline (week 1)
Quote requests received186
Quotes answered (rep-approved)171~140 TYPICAL
Escalated to reps (correctly out of scope)15
Median response time (draft → approved → sent)11 MIN4 HR 05 MIN
After-hours requests answered next-AM queue23MOSTLY UNANSWERED
Loads under track & trace94
Check-ins completed without a phone call86 (91%)0
Exceptions escalated8
Approved-without-edit rate81% ↑79% (WK 3 SHADOW)

Dollars · illustrative model

After-hours quotes that converted to covered loads6
Est. margin on those loads (your avg $250/load)$1,500
Rep hours not spent on check calls (~12 min ea × 86)~17 HRS

ILLUSTRATIVE MODEL — MARGIN ATTRIBUTION USES YOUR TMS NUMBERS, FLAGGED WHERE CAUSALITY IS SOFT.


Exceptions caught this week

TUE 14:22 CARRIER SILENT 6+ HRS ON HOT LOAD → REP PAGED, RECOVERED WED 09:51 LOCATION MISMATCH: CHECK-IN SAID I-75, ELD LANE SAID I-24 → FLAGGED THU 16:08 ⚑ REMIT-TO EMAIL CHANGED MID-LOAD + NEW MC AGE < 6 MO → DOUBLE-BROKER PATTERN · ESCALATED · YOUR CALL, NOT OURS FRI 11:30 QUOTE REQUEST FROM UNKNOWN SHIPPER DOMAIN → HELD FOR REP (POLICY)

Billing · included items

Items processed (quotes + check-ins)280 / WK · ~1,120 MO PACE
Plan includes500 ITEMS / MO
Projected overage @ $1.50/item~620 × $1.50 = $930

OVERAGE IS SHOWN WEEKLY SO THE INVOICE IS NEVER A SURPRISE.


Maintenance notes

Nightly regression: 7/7 green nights. One parser patch (carrier reply format) deployed Thursday, re-tested same day. No action needed from you.

STATUS: RUNG 0 — EVERY SEND REP-APPROVED. GO/NO-GO MEETING SCHEDULED DAY 30.

FORM DD-07 · WEEKLY REPORT · MONTH 4SAMPLE REPORT

Summary · week of (sample)

MetricThis weekTrend
Quote requests received214
Quotes answered203
  · rep-approved sends141
  · auto-sent on your 3 designated routine lanes (Rung 2)62YOU SIGNED THESE LANES
Median response time (auto-sent lanes)96 SEC
Median response time (gated lanes)9 MIN
After-hours requests answered31 (14 AUTO ON SIGNED LANES)
Loads under track & trace128
Check-ins completed without a call119 (93%)
Exceptions escalated9
Approved-without-edit rate (gated lanes)91%
Auto-send accuracy audit (weekly spot-check, 20 sampled)20/20 WITHIN RULESHARD GATE

Dollars · illustrative model

After-hours quotes converted to covered loads11
Est. margin on those loads$2,750
Rep hours returned (check calls + drafting)~26 HRS

ILLUSTRATIVE MODEL — SAME CAVEATS, EVERY WEEK, IN WRITING.


Exceptions caught this week

MON 03:12 CARRIER ETA SLIP DETECTED FROM SMS REPLY → CONSIGNEE NOTIFIED 6 AM WED 13:40 ⚑ MC NUMBER ON CHECK-IN ≠ MC ON RATE CON → DOUBLE-BROKER FLAG → LOAD FROZEN IN QUEUE · YOUR DISPATCH DECIDED IN 11 MIN FRI 17:55 NEW SHIPPER REQUEST ON A SIGNED AUTO-LANE → STILL HELD FOR REP (NEW SHIPPERS NEVER AUTO-SEND, REGARDLESS OF LANE)

Billing · included items

Items processed331 / WK · ~1,320 MO PACE
Plan includes500 ITEMS / MO
Projected overage @ $1.50/item~820 × $1.50 = $1,230

Maintenance notes

Rate feed schema change (vendor-side) absorbed Tuesday — caught by the 2 AM suite, patched before your morning. Rung-2 lane audit: all clear. Your standing right to revoke auto-send is one email away.

STATUS: RUNG 2 ON 3 LANES · RUNG 0/1 EVERYWHERE ELSE · FRAUD PATHS ALWAYS HUMAN.

The retainer

The anti-silent-failure contract.

The $2,000/month isn't a license fee — it's a maintenance crew. Email formats change, TMS vendors ship breaking API versions, models get upgraded. All of it is our problem, on our clock, covered by the retainer. Here's exactly what that means.

FORM DD-08 · MAINTENANCE COVERAGERETAINER TERMS — DRAFT, FINALIZED IN PILOT AGREEMENT
When this happensWhat we doTarget
A carrier changes their email or SMS reply formatNightly regression flags it; affected pattern routes to humans (no silent guessing); parser patched and re-testedNEXT BUSINESS DAY
Your TMS ships a breaking API versionConnector updated on our side; sandbox-verified before it touches live loadsPRIORITY · ON US
DAT / Greenscreens changes a feed schemaAbsorbed in the integration layer; you never see it except as a line in the weekly reportSAME WEEK
The underlying model gets upgradedFull regression re-baseline against your corpus before any swap; rollback if scores dipNO SWAP WITHOUT GREEN SUITE
You change your margin rulesRules updated, re-tested against history, deployed — with a diff in the weekly report2 BUSINESS DAYS
Processed-item volume drops abnormally (the classic silent failure)Volume-anomaly alarm pages us; we call you — not the other way aroundALARM < 1 HR
Anything breaks at 2 AMMonitoring pages the founder. Your reps find out from us, with a fix timeline, in the morningPAGED, NOT POSTED

ALL OF THE ABOVE IS INSIDE $2,000/MO (INCL. 500 PROCESSED ITEMS; $1.50/ITEM AFTER). MAINTENANCE IS NOT AN UPSELL. MAINTENANCE IS THE PRODUCT.


The "no" list

What we refuse to do.

A narrow service stays reliable by staying narrow. These are the things we turn down — including the ones that would be easy money.

Refused

Unsupported TMS platforms

Aljex, Tai, McLeod, Turvo, Ascend. That's the list. A connector we can't regression-test nightly is a silent-failure factory, so we don't wing it — even for a signed check.

Refused

Full autonomy on day one

Some prospects ask for it. The answer is no, even when it costs us the deal. The trust ladder exists because your customers can't tell our mistake from yours.

Refused

Voice agents

Phone calls are harder to parse, harder to test, and harder to audit than email and SMS. We don't ship what we can't put through the nightly suite.

Refused

Multi-vertical sprawl

No "and also insurance, and also recruiting." Freight brokerage email, two agents, done well. Depth is the moat; sprawl is how option B happens.

SAYING NO IS CHEAPER THAN CHURN — FOR BOTH OF US.

Fit check

Does your desk qualify?

Thirty seconds, three questions, an honest answer. We qualify hard because a bad-fit pilot wastes your $2,500 and our month.

FORM DD-09 · FIT CHECKNO EMAIL REQUIRED
RESULT APPEARS BELOW — NOTHING IS SENT ANYWHERE.
Inspected · gated · maintained

Thirty days. One inbox. Hard numbers decide.

$2,500 pilot, credited to setup if you continue. If the before/after sheet doesn't earn the next month, walk — we'll hand you the data on the way out.

NO AUTONOMOUS SENDING ON DAY ONE. NO EXCEPTIONS — SEE THE "NO" LIST ABOVE.