J.S.Cock · AI Inquiry Demo
← All inquiries · Inquiry detail view
End-to-end flow

JT123478

One inquiry, all stages. Each box below is the actual artifact persisted by the pipeline — not a mockup. The brief asks: "does Kine actually work?". This page is the answer.

9 stages
  1. 1

    Email arrives in M-Files inbox

    Step 01

    Inquiry stored at .tmp/jscock_mfiles/JT123478/inquiry.eml

    ok
    From: orjanjoha@hotmail.com
    Subject: 14 toms spade ventil/stafsjø ventil med luftsylinder. E d nåkka dokker kan skaffe og til hvilken pris
    Body: 617 chars
    Attachments: image001.jpg, image704858.gif, image666111.png
    Show full email body
  2. 2

    AI extracts structured JSON

    Step 02

    1 item(s) detected · language=no · intent=new_inquiry

    ok
    Norwegiannew inquiry1 item(s)
    Show extracted JSON
    {
      "language": "no",
      "intent": "new_inquiry",
      "customer": {
        "company": null,
        "contact_name": "Ørjan Johannessen",
        "contact_email": "orjanjoha@hotmail.com",
        "phone": null,
        "vat_number": null,
        "address": null
      },
      "items": [
        {
          "line_no": 1,
          "quantity": null,
          "unit": null,
          "product_reference": "spade ventil/stafsjø ventil med luftsylinder (spade valve/stafsjø valve with air cylinder)",
          "manufacturer": null,
          "manufacturer_part_no": null,
          "cross_ref_to": null,
          "specs": {
            "dimension": "14 toms",
            "pressure_rating": null,
            "material_body": null,
            "material_seal": null,
            "actuation": "pneumatic",
            "voltage": null,
            "torque_or_flow": null,
            "temperature": null,
            "thread_or_connection": null,
            "other": []
          },
          "missing_fields": [
            "quantity",
            "pressure_rating",
            "material_body",
            "material_seal",
            "temperature",
            "thread_or_connection",
            "manufacturer",
            "manufacturer_part_no"
          ]
        }
      ],
      "application_context": null,
      "attachments_referenced": [],
      "delivery": {
        "location": null,
        "deadline": null,
        "incoterms": null
      },
      "previous_reference": {
        "is_update_of": null,
        "notes": null
      },
      "completeness": "partial",
      "missing": [
        "quantity",
        "delivery location",
        "delivery deadline",
        "application context",
        "complete product specifications"
      ]
    }
  3. 3

    Completeness verdict: partial

    Step 03

    Missing: quantity, delivery location, delivery deadline, application context, complete product specifications

    ok
    partial
    • quantity
    • delivery location
    • delivery deadline
    • application context
    • complete product specifications
    Rule trace (deterministic)
    Verdict: partial · 1 of 1 items need clarification
    • item 1: × needs follow-upno quantity
  4. 4

    AI drafts follow-up email to customer

    Step 04

    Drafted in no, awaiting human approval.

    ok
  5. 5

    Customer creation deferred

    Steps 08 + 09

    Customer company name not extracted — needs human review.

    skip
  6. 6

    Match items against the J.S.Cock catalog

    Steps 12 + 13

    Top-5 candidates per item via TF-IDF over 18.8k SKUs (see Matches tab on /inquiry/JT123478).

    ok
  7. 7

    Score supplier quotes

    Steps 16 + 17

    No supplier replies received yet — scoring waits for at least one.

    skip
  8. 8

    Draft quote in Business Central

    Steps 10 + 20

    Waiting for the supplier quote to land before drafting.

    skip
  9. 9

    Human approval (sales engineer)

    Step 21

    In production this becomes a real review UI in Week 2-3. Demo: approve / edit / reject from the page.

    pending

Process map: see how this aligns with JS Cock’s diagram →