Main site

Transportation AI agent · Load matching

Load-Matching and Tender AI Agent

Shortlist carriers against your own written qualification standard, attach the evidence, and hold the tender for a named person — selecting the carrier is now the act that gets sued.

Typical delivery
4–6 weeks
Deployment
Your stack
Person tenders
Evidence first
After launch
Agent Care

What this agent does

Shortlists the carrier, never tenders the load

In

  1. 01

    Ingest the load, its lane requirements and your written carrier standard from supported sources.

  2. 02

    Re-verify active operating authority at the moment of tender, not only at onboarding, and keep the timestamp.

Reason

  1. 03

    Confirm insurance with the insurer or the agent, rather than from a certificate the carrier supplied.

  2. 04

    Check the limits actually carried against what this load requires, and record the channel used.

  3. 05

    Rank the shortlist against your written standard, and mark what the record does not establish.

Decide

  1. 06

    Log every override of that standard with the person who made it and the reason they gave.

  2. 07

    Route the shortlist to a named person, who decides which carrier is tendered.

Out

  1. 08

    Retain what was checked, what was found and who decided, for the period the records rule requires.

  2. 09

    Execute write actions only inside the approval boundaries agreed during implementation.

Product statement

The agent ranks and evidences; a named person selects and tenders, and your firm stays the broker of record.

Example workflow

One load, request to tender

  1. 1Load receivedShipper order, TMS load record, board posting or lane commitment
  2. 2Evidence gatheredOperating authority, insurance confirmed with the insurer, equipment and lane history, each with its source and time
  3. 3Shortlist preparedRanked carriers, evidence and confidence
  4. 4Controls appliedAuthority re-checks at tender, insurer-confirmed cover, your written standard and the confidence threshold

No human action required

Stages 1 to 4 run unaided and tender nothing — the shortlist is assembled, and the person's lane opens at the confidence gate.

5DecisionBranches at the confidence threshold
High confidence

Goes to the named person to tender.

Low confidence

Adds a compliance read first.

A person tenders

The shortlist is held with its evidence, its gaps and the confidence.

Tender · Reorder · Send to compliance

Tendered — offer out to the carrier

  1. 6Load records updatedOnly where write access and approval policy allow it
  2. 7Outcome evaluatedEvidence gaps, overrides recorded, tender acceptance and what each selection actually rested on

Overrides

Every override of the standard is counted in the evaluation.

What should not run autonomously

Human approval stays in control

Outside the boundary — human approval required

8 items

  • Tendering a load or issuing an offer to a carrier.
    • Deciding which carrier is good enough to be used.
    • Waiving your written carrier-qualification standard.
    • Accepting a certificate the carrier supplied itself.

  • Automation boundary

    Agent acts unaided

    • ✓Gather authority, insurance and lane evidence for each.
    • ✓Rank the shortlist against the standard you have written.
    • ✓Mark what the record leaves unestablished, carrier by carrier.
    • ✓Hold the shortlist for the named person who tenders the load.

    Any write happens inside the boundaries agreed at implementation, never ahead of approval.

    • Directing the route, the driver or the manner of haul.
    • Describing your firm's operations as a carrier's.
    • Deciding whether a transparency waiver still holds.
  • Changes to the qualification standard or its rules.

Example output

One shortlist, annotated

Automating selection concentrates the exposure and records it — so each ranking carries its evidence.

Matching output · single load

Illustrative example
Load
Reefer, long haul
Top of the shortlist
Ranked first on lane history and equipment, with one gap noted
Authority re-checked
11:42, at tender
Insurance confirmed
Confirmed with the insurer direct
Confidence
89%
Standard applied
Your written qualification standard

As received

Taken from the FMCSA record and the insurer's own confirmation.

What was checked

  • Active operating authority
  • Insurance, with the insurer
  • Limits against this load

Why it is first

It meets the standard you wrote — no court set that standard.

What the score decides

Below the configured threshold the shortlist picks up a compliance read before.

Action

TenderReorderSend to compliance

Value

Where AI adds value

The same four claims, placed at the point in the workflow where each one applies.

Where the value lands

Value 01 – 04

Every load

From the TMS or the load queue

03Matching

Shortlist against your standard

Draw on the authority record, insurer-confirmed cover and the lane history you already hold.

01Approved path

Shortlist it, a person tenders

Routine lanes come back ranked with the evidence attached.

02Human review

Pass the rest to compliance

Evidence gaps, overrides and low-confidence shortlists are marked, so the desk's read starts where risk concentrates.

04

Build an evidence trail

The match, the evidence behind it and the person who tendered stay on the load file.

Integrations

Typical integrations

Five system groups connect to the same agent. Which of them are in scope is decided in discovery.

  • TMS and load records

    • McLeod
    • MercuryGate
    • Trimble
    • Revenova

  • Load boards

    • DAT
    • Truckstop
    • Private capacity networks

  • Authority and insurance

    • FMCSA records
    • Motus
    • Insurer and agent confirmation

Agent

Load matching and tender

  • Gathers the evidence
  • Ranks the shortlist
  • Holds for tender

  • Onboarding and files

    • Carrier onboarding tools
    • Document archives

  • Observability & evaluation

    • OpenTelemetry
    • Langfuse
    • Supported monitoring/evaluation sources

Integration availability depends on the client's existing systems and API access.

Agent controls

Six layers between the model and the tender

The layers sit one inside the next. What none of them catches is in the map below.

L6 · Outermost — last line of defence

Inward → L1 · closest to the model

L6Rollback / safe modeFall back to listing available capacity when evaluation or production signals degrade.Roll back
L5Version monitoringTrack model, prompt, standard and check-configuration changes.Track
L4TraceabilityRecord what was checked, what came back and when it was checked.Record
L3Tender heldHold the shortlist for the named person; it governs who tenders, not whether the carrier is sound.Gate
L2Qualification rulesTest each carrier against your written standard; a failure returns the shortlist with the reason.Restrict
L1Confidence thresholdsRoute low-confidence shortlists to a compliance read before anyone sees them.Require review
Model coreShortlist prepared — the ranking, the evidence, the gaps and confidence
L1 – L2
Test whether a carrier may be put forward
L3
Puts the tender in a person's hands
L4 – L5
Hold the match and the evidence behind it
L6
Falls back to listing available capacity when signals degrade

How Nestack evaluates it

Evaluate the whole selection — not only the carrier that took the load.

Coverage runs the whole depth of the workflow, and every layer is cut by slice.

Surface — the shortlist the desk reads

  1. E1Final-output evaluationWas each check re-run at tender rather than carried from onboarding?
  2. E2Step-level evaluationDid the agent apply your written standard, in its current version?
  3. E3Tool evaluationDid it read the correct authority record and write the correct load?
  4. E4Confidence calibrationDo low-confidence shortlists actually attract more reordering?
  5. E5Slice evaluationHow does performance change across specific carrier segments?
  6. E6Business outcomeHow many shortlists were reordered, and how many should have been?

Floor — the selection your firm answers for

Failure modes

Where each failure originates in the agent

Seven ways a shortlist goes wrong, by stage.

Agent lifecycle

Direction of processing →

  1. 01 · Retrieval

    1 mode

    • AQ-03

      Onboarding check reused

      A months-old authority result stands in at tender.

    Stage gathers

    Authority, insurance, equipment and lane history

  2. 02 · Reasoning

    2 modes

    • AQ-04

      Standard applied loosely

      A qualification rule is read as a preference, not a bar.

    • AQ-06

      Ranking read as decision

      The order of the list is treated as the selection itself.

    Stage proposes

    The ranking, the evidence and confidence

  3. 03 · Tool / write

    2 modes

    • AQ-02

      Tender sent without a person

      An offer reaches a carrier before anyone has selected.

    • AQ-05

      Duplicate tender

      The same load is offered twice into one lane.

    Stage writes

    Only where write access and approval policy allow it

  4. 04 · Output

    1 mode

    • AQ-01

      Carrier-supplied certificate

      Cover is taken from a document the carrier provided.

    Stage returns

    The shortlist the desk reads and tenders from

  5. 05 · Change / Version

    1 mode

    • AQ-07

      Silent standard regression

      A model or rule change relaxes what the shortlist admits.

    Stage tracks

    Model, prompt, qualification standard and checks

  • Sev-1 · tendered with no person
  • Sev-2 · put forward against the standard
  • Sev-3 · evidence degrades, list routes to review
See the complete Transportation failure-mode catalogue

Affected slices

One segment carries the selections you would regret

Carriers new to your file is the segment that hides — small enough to disappear into a total, and carrying most of the reordering. Nestack reports the reorder rate by carrier segment, not only in total.

Slice performance — reported separately, not only in aggregate

Illustrative example
SliceFailure rateLiftLift vs. thresholdStatus
Carriers new to your file6.6%3.2×Review
Brokered spot coverage4.5%2.2×Review
Lanes covered under time pressure3.9%1.9×Watch
Contracted core carriers1.2%0.6×Normal

Bar: reorder-rate lift vs. contracted-core-carrier baseline · scale 0–4.0× · tick marks the 2.0× review threshold

2 of 4 slices over threshold

Evidence-linked improvement

Write the case, then close the cycle

The cycle ends in a regression case, not in a meeting about what happened. That suite is what the next tender put out to a carrier is measured against.

Improvement cycle · five stages

Switchback — the path turns at Improve and returns at Learn

  1. 01

    Detect

    Reorder rate rises in a carrier segment.

  2. 02

    Diagnose

    Take the shortlist apart: which check produced the rank, and which one should have blocked it.

  3. 03

    Improve

    Every change is versioned against the tenders that exposed it.

  4. 04

    Verify

    A failing case holds the release back.

  5. 05

    Learn

    The case joins the suite for good, and the matching rules are revisited.

  6. Learn → Detect

    The return edge. The next detection runs against a suite one case longer.

Typical build scope

Twelve workstreams across six weeks

The build scope read against the delivery timeline. Week structure follows the six-week plan — discovery, sources, matching workflow, evaluation, integration, then production validation and handover.

Workstream

Week 1Week 2Week 3Week 4Week 5Week 6

  1. 01Matching and tender workflow discovery and boundaries.Week 1
  2. 02TMS and load-board assessment.Week 1, Week 2
  3. 03Carrier-qualification standard and evidence mapping.Week 1, Week 2
  4. 04Load and carrier-record ingestion.Week 2
  5. 05Ranking logic and evidence binding.Week 2, Week 3
  6. 06Confidence scoring and compliance routing.Week 3
  7. 07Tender approval workflow.Week 3, Week 4
  8. 08TMS and load-board integration.Week 5
  9. 09Selection-evidence cases.Week 4, Week 5
  10. 10Guardrails and tender controls.Week 4
  11. 11Match-trail instrumentation.Week 5, Week 6
  12. 12Deployment, documentation and Agent Care handover.Week 6

12 workstreams · 6 weeks · bar shows the weeks a workstream is active — several run in parallel

Final scope and sequence confirmed in discovery

Engagement tiers

What each tier includes

Rows are the capabilities named in each tier's scope. Higher tiers include everything below them.

Capability✓ in scope · — not at this tierPilotOne lane set, one officeProductionProduction TMS and load boardsAdvancedMultiple entities / regions
Introduced at Pilot
Shortlisting to your written standard✓✓✓
Tender held for a person✓✓✓
Match-quality baseline✓✓✓
Introduced at Production
Reporting by lane and equipment—✓✓
Tender workflow in your systems—✓✓
Approved write-back—✓✓
Load-board integration—✓✓
Introduced at Advanced
Multi-entity qualification rules——✓
Multi-stage compliance approvals——✓
High tender volume——✓
Multi-entity tender controls——✓
Build priceFrom $5,000From $8,000Custom quote
Final build price
Confirmed after discovery based on integrations, workflow complexity, transaction volume, approval controls and deployment requirements.
Separate from build
Build pricing is separate from recurring Agent Care, which covers managed monitoring, evaluations, incidents and verified improvements after launch.

What we need from you

What you bring, and what we build with it

Each input maps to a piece of build scope and a week in the delivery timeline.

  1. 01Your written carrier-qualification standard

    Load and carrier-record ingestion and evidence mapping

    Week 1

  2. 02Representative past selections

    Matching baseline, ranking logic and evidence binding

    Week 2

  3. 03Your insurer and authority check routes

    Carrier-qualification standard and evidence mapping

    Week 1

  4. 04Access to relevant APIs, feeds or exports

    TMS and load-board assessment, then integration setup

    Week 2

  5. 05Matches you would not want tendered

    Evidence cases and the evaluation suite

    Week 4

  6. 06What must reach a person before a load is tendered

    Confidence scoring, compliance routing, guardrails and tender controls

    Week 3

  7. 07Named people who tender the load

    Tender approval workflow, then pilot and production validation

    Weeks 5–6

Nothing else is required

Deployment, documentation and Agent Care handover are ours.

Delivery timeline

Four phases across six weeks

The bands follow the real work, which is why evaluation and pilot share the fifth week.

Phase

W1

W2

W3

W4

W5

W6

  1. DiscoveryW1
  2. BuildW2 – W3
  3. EvaluateW4 – W5
  4. Pilot & LaunchW5 – W6

Week focus

W1Matching workflow discovery, standard mapping and the boundary

W2Load-board and authority integration and the matching baseline

W3Matching workflow, confidence logic and tender controls

W4Evaluation suite, evidence cases and failure-mode testing

W5TMS integration, a pilot lane set and targeted corrections

W6One tender cycle worked under the desk, then handover

Reading the band
Each bar covers only the weeks its work is named in. The week 5 overlap is real, not padding.
At the end of W6
The cycle closes validation and Agent Care assumes monitoring.
Duration
Six-week plan shown · typical delivery 4–6 weeks depending on scope confirmed in discovery.

Next step · Transportation AI agent

Build a matching agent that shortlists and leaves the tender to a person.

Show us your lanes, your boards and how a load gets covered today. Bring the written qualification standard and the insurer verification route — without a written standard there is nothing for the agent to enforce and nothing for you to show later.

Typical delivery
4–6 weeks
Deployment
Your TMS and boards
Person tenders
Evidence first
After launch
Agent Care available

Nestack Agents · Load matching and tender

AGT-TR-08 · Agent Care available after launch

Build a load-matching agent

Four questions so the call starts from your actual setup — then pick a time that suits you.

About you
Your loads

No obligation. We’ll map the workflow, name the automation boundary and tell you if an agent is the wrong answer.