DEADHEADDESK · FORM DD-00 · REV A THE DESK IS ON · --:--:-- LOCAL
Agent 01 · Form DD-01 · Spot-quote desk

The Spot-Quote Agent

It reads the request — typos and all. It pulls your lane history from the TMS and a live market rate from DAT or Greenscreens. It prices under your margin rules and drafts the reply. Then it stops, and waits for a human. Every quote, every time, until you say otherwise.

READS: EMAIL · TMS · DAT/GREENSCREENS  ·  DRAFTS: ~90 SEC (DESIGN TARGET)  ·  AUTO-SENDS: NEVER ON DAY ONE

EVERY QUOTE · DAY ONETHE GATE

Draft · held for rep approval

KC, MO → Memphis, TN · 53' dry van $1,460 all-in · picks Wed AM 2 confirmations flagged in reply


THE AGENT CANNOT SKIP THIS STEP. ONLY YOU CAN WIDEN IT — SEE THE TRUST LADDER BELOW. SAMPLE DATA.

The routing slip

Inbox to logged quote: nine stations, one gate

Click any station to see what actually happens there. Station 08 is the one that matters: a human, with a button.

01Inbound emailT+0S
WATCHES quotes@yourbrokerage.com 24/7. The agent reads a copy — nothing is moved, deleted, or replied to from the inbox itself.
02ClassifyT+2S
QUOTE REQUEST? Scored 0–1. Spam, carrier check-ins, and rate-con noise are filtered out — and not billed as processed items.
03ExtractT+6S
ORIGIN · DEST · EQUIPMENT · WEIGHT · DATES — each field gets its own confidence score. Weak fields are flagged so the reply asks instead of guesses. Too weak overall → straight to a human.
04Lane historyT+18S
READ-ONLY TMS QUERY: your past loads on the lane — count, rate band, customer record. Works with Aljex, Tai, McLeod, Turvo, Ascend.
05Market rateT+24S
DAT / GREENSCREENS: live spot rate + 7-day trend. The exact numbers used are snapshotted into the audit trail — you can always see what the market said at that minute.
06Margin rulesT+27S
YOUR floors and targets, by lane / customer / equipment — a versioned ruleset you signed off on. Below-floor math is never drafted; it escalates instead of discounting.
07DraftT+41S
A priced reply in your house voice. Anything the agent assumed or couldn't read appears in the draft as a question to the shipper — in writing, on the record.
08Rep approvalThe gateHOLDS
THE DRAFT STOPS HERE until a rep clicks approve, edit, or reject. Day one this is 100% of quotes. The agent cannot open this gate — only you can widen it, lane by lane, on the trust ladder below.
09TMS logON VERDICT
Quote record written back to the TMS with the full audit trail: source email, extraction, rate snapshot, rule version, draft, verdict, approver, timestamps.

STATION TIMES ARE DESIGN TARGETS FROM OUR TEST CORPUS, NOT GUARANTEES. THE GATE HAS NO TIMER — IT WAITS FOR YOUR REP.


The hard part is the reading

Watch it work a messy one

Real quote requests come in misspelled, half-specified, and sent from a phone at a loading dock. Here's how the agent reads one — and what it does about the parts it can't read. Click any highlighted phrase or table row to trace it.

ANNOTATED WALKTHROUGHSample data

Inbound · Tue 2:14 PM · quotes@yourbrokerage.com

From: shipping@orchardstatefoods.com Subject: RATE?? hey can u quote a van load kanas city to memhpis 26 plts of grocery, dock high needs to pick up tomorow morning how fast can you cover? lmk sent from my iPhone

GREEN = READ CLEAN  ·  ORANGE = FLAGGED — THE REPLY WILL ASK

Extraction · per-field confidence

FieldRead asConf
ORIGINKANSAS CITY, MO ("kanas city" — typo resolved)0.94
DESTMEMPHIS, TN ("memhpis" — typo resolved)0.95
EQUIP⚑153' DRY VAN — ASSUMED FROM "van"0.82
WEIGHT⚑2NOT GIVEN — 26 PALLETS GROCERY
PICKUPWED 06/11, AM ("tomorow" resolved vs. received-date)0.91

Two fields didn't clear the bar. The agent doesn't silently guess on them — it prices what it can defend and puts both flags in the reply as questions.


Draft reply · held for rep approval · both flags confirmed in writing

Hi — got it. Kansas City, MO → Memphis, TN, picking up tomorrow (Wed 6/11) morning: $1,460 all-in, quoting as a 53' dry van — say the word if you need something else ⚑1. Rate is good for standard grocery weight, under 42,000 lbs — can you confirm total weight on the 26 pallets? ⚑2 We can commit capacity this afternoon. Reply here and we'll lock it. — Drafted by DeadheadDesk · held for rep approval

Why this matters

A wrong assumption that's stated costs you an email. A wrong assumption that's hidden costs you a claim. Every flag the extractor raises becomes a sentence in the draft — so the shipper confirms it, in the thread, before anyone books a truck.

And if too much is unreadable — no lane, no date, gibberish — the agent doesn't draft at all. It escalates with whatever it managed to parse. See the escalation list below.

RATE SHOWN IS SAMPLE DATA. CONFIDENCE THRESHOLDS ARE TUNED PER CLIENT DURING THE PILOT.

Autonomy is earned, not assumed

The trust ladder. You hold it.

Nothing auto-sends on day one — and nothing ever does unless you switch it on. The agent earns rungs with a scorecard; only you grant them. Drag the slider.

TRUST LADDER · CLIENT-CONTROLLEDRUNG 0 OF 3
Day 1Weeks 2–5Week 6+Months in

Rung 0 — A rep approves everything


Held for approval
 
100%
Auto-sent
 
0%

SPLITS SHOWN ARE ILLUSTRATIVE — YOUR WHITELIST AND CAPS DECIDE THE REAL NUMBER. ONE TOGGLE DROPS THE WHOLE DESK BACK TO RUNG 0, INSTANTLY.


Honest limits

What always goes to a human

The agent's most important skill is knowing what it doesn't know. This list never auto-quotes — at any rung of the ladder, forever. That's not a limitation we apologize for; it's most of why the desk is safe to run.

STANDING ESCALATION ORDERSNO EXCEPTIONS, NO LADDER RUNG OVERRIDES
What came inWhy it never auto-quotesWhat your rep gets
Multi-stop loadsStop-off fees and routing order are judgment calls, not lookups.FULL EXTRACTION + LANE DATA, PACKAGED
Reefer partialsTemp liability plus consolidation math — humans only.TEMP, PALLET COUNT, COMPATIBLE-FREIGHT NOTES
HazmatPlacards, endorsements, insurance. Not guessing territory.FLAGGED CLASS + HANDLING REQUIREMENTS
Brand-new shippersNo lane history, no credit signal, no pattern to price against.LOOKUP RESULTS + NEW-SHIPPER CHECKLIST
Low-confidence extractionIf the agent isn't sure what it read, it doesn't price it.RAW EMAIL + WHAT IT COULD AND COULDN'T PARSE
Anything that smells like fraudMismatched reply-to, look-alike domains, rates too good to be real.RED-FLAG REPORT — BEFORE ANYONE REPLIES

ESCALATIONS COUNT AS PROCESSED ITEMS — THE AGENT DID THE READING AND THE PACKAGING. WHAT IT DIDN'T DO IS GUESS.

Plumbing

What it reads. What it writes back.

Lane history comes out of your TMS through a read-only, credentialed connection scoped during setup week. Everything the agent does goes back in as a record your auditors — and your gut — can check.

READS · READ-ONLYSCOPED AT SETUP
Lane historyPAST LOADS · RATE BAND · CARRIERS
Customer recordTERMS · CONTACTS · NOTES
Margin rulesetYOUR FLOORS/TARGETS · VERSIONED
Market ratesDAT · GREENSCREENS

ALJEXTAIMcLEODTURVOASCEND

DIFFERENT TMS? CHECK FIT — WE SAY NO WHEN IT WON'T WORK.

WRITES BACK · EVERY QUOTESample record
QUOTE-2026-0610-114 SRC_EMAIL ... msg_8c41 (quotes@) rcvd 14:12:08 EXTRACTION .. KCMO→MEM · DV53 (0.82, flagged) wt UNCONFIRMED (flagged) LANE_HIST ... 9 loads · $1,310–$1,495 · read 14:12:31 MARKET ...... DAT $1,512 · GS $1,488 · snap 14:12:33 RULES ....... floor 12% · ruleset v7 (signed 05/28) DRAFT ....... $1,460 all-in · 2 confirmations GATE ........ HELD → approved D. REYES 14:19:42 (rep edit: +fuel-surcharge note) TMS_WRITE ... quote logged · thread linked

When a shipper disputes a number in October, you pull this record and see exactly what the market said, which ruleset priced it, and which human approved it — in one place.

The harness

A shipper changes their email format. Then what?

Every DeadheadDesk agent runs inside an eval and monitoring harness. The failure mode is designed: a degraded agent escalates more. It never guesses more.

Nightly · regression

The agent re-runs a corpus of real freight quote emails — clean ones, messy ones, the ones that broke it before — scored per field against known answers. A score dip blocks changes from shipping.

Always on · drift watch

Confidence and rep-edit rates are tracked per sender. When a shipper's new ERP changes how their requests read, confidence sags — and the harness notices before a human would.

The failsafe

A low-confidence email doesn't get a worse quote. It gets no quote — it routes to your reps as if the agent were never there, and the alarm pages us. Not you.

The patch

We fix the parser, add the new format to the regression corpus so it can't quietly break again, and you read about it in the weekly report — in dollars and loads, not log files.

SILENT-FAILURE ALERTING IS A CONTRACT TERM, NOT A FEATURE FLAG. IF THE DESK DEGRADES AND YOU FIND OUT FROM A SHIPPER BEFORE YOU HEAR IT FROM US, THAT'S OUR FAILURE — IN WRITING.

PROJECTED OUTCOMESSample data — pre-launch

0 SEC

Median request → draft

Design target from our test corpus.

0%

After-hours capture

Overnight requests with a draft waiting before your reps sit down.

+0%

Quote-to-cover

Illustrative model: faster replies win more of the loads you quote.


THESE ARE DESIGN TARGETS AND MODEL OUTPUTS — NOT CUSTOMER RESULTS. THE 30-DAY PILOT REPLACES THIS STRIP WITH YOUR OWN BEFORE/AFTER NUMBERS, MEASURED ON YOUR INBOX.

Quoting FAQ

The questions brokers actually ask

Who's liable if a bad quote goes out?
Until you grant auto-send, no quote leaves without a rep's click — so a sent quote is owned the way it is today: by your desk, with better paperwork behind it. If you later enable auto-send and a quote misprices inside the guardrails you set, responsibility is spelled out in plain terms in the service agreement before anything goes live. Either way, every quote carries the full audit trail — what was read, what the market said, which ruleset priced it, who approved it. Ask for the agreement on the pilot call; it's short.
How do margin rules get set?
In discovery week we sit with whoever owns pricing and encode your floors and targets — by lane, customer, and equipment — into a versioned ruleset you sign off on. The agent cannot quote below floor: below-floor math escalates instead of discounting. Rule changes are a ticket to us with same-day turnaround, and every quote logs which ruleset version priced it.
What happens with weird requests?
They escalate — fast, and with the reading already done. Multi-stop, reefer partials, hazmat, new shippers, anything the extractor can't read with confidence, anything that smells like fraud: the agent packages what it parsed and hands it to your rep instead of guessing. The standing escalation list on this page never auto-quotes at any rung of the trust ladder.
What counts as a "processed item"?
One inbound email the agent works end to end: classified, extracted, and either drafted or escalated with context. Spam, duplicates, and non-quote mail the classifier filters out are not billed. 500 items a month are included in the $2,000 retainer; $1.50 per item after that. The weekly report shows the count, itemized — no surprise invoices.
Can we turn auto-send off?
Yes — instantly, per lane or globally. One toggle drops the whole desk back to day-one mode where a rep approves everything. Auto-send is opt-in and never the default: day one nothing auto-sends, and if you never touch the ladder, nothing ever does.
The gate stays shut until you open it

Put the quote desk on one inbox.

30-day paid pilot, $2,500, credited against setup. Shadow mode first, approval gates always, and a go/no-go decision made on hard before/after numbers — not a sales deck.

$2,500 PILOT (CREDITED) · $7,500 SETUP · $2,000/MO INCL. 500 ITEMS · $1.50/ITEM AFTER