AI Rate-Con Parsing Inside X-TMS
Stop typing rate confirmations. Drop them into X-TMS — your TMS does the work in seconds.
Rate-con typing is the slowest part of dispatch — X-TMS removes it
Manual rate-con entry takes 5–10 minutes per load. Multiply by 30 loads a day and that's three hours of typing. Typos creep in. The wrong rate gets entered. The wrong pickup date gets confirmed.
X-TMS lets you drop the PDF — or a stack of PDFs — into the upload box. AI extracts every field, validates against historical patterns, and creates ready-to-dispatch loads instantly.
How it works
-
1
Drag and drop the rate confirmation
PDF, PNG, JPG — single file or batch up to 50. Drag-and-drop or click to upload.
-
2
AI parses each document
Origin, destination, rate, total miles, weight, commodity, equipment type, pickup and delivery windows, reference numbers, broker contact, and any special instructions.
-
3
Confidence-scored field extraction
Each extracted field gets a confidence score. Fields below threshold are flagged for dispatcher review.
-
4
Loads created in your dispatch queue
Within seconds. Original document attached. Ready to assign.
-
5
Historical rate validation
X-TMS cross-checks the offered rate against your historical lane data — alerting if it's below typical for that route.
By the numbers
What's included
- PDF parsing (multi-page supported)
- Image parsing (PNG, JPG, HEIC, scanned documents)
- Plain text parsing (paste from email)
- Batch upload up to 50 documents at once
- Per-field confidence scoring
- Historical lane rate validation
- Automatic load creation in X-TMS
- Original document attached to load for audit
- Multi-language rate confirmation support
- API endpoint for programmatic upload from external systems
- Drag-and-drop UI plus mobile camera capture
- Field-level edit before commit — fix anything the AI got wrong
- Smart duplicate detection — flags potential duplicates of existing loads
- Broker contact extraction with automatic customer record matching
Best fit for
Compress an hour of typing into 5 minutes of review.
Same engine works for outbound rate cons to your carriers.
Process a stack of overnight emails in one batch.
Bulk-import historical rate cons during onboarding.
What to expect during rollout
Most teams underestimate the people-and-process work that surrounds any new technology. AI Rate Con Upload is technically straightforward to switch on, but a smooth rollout still benefits from a short coordinated effort across dispatch, IT, and ownership. Below is what we typically see in successful deployments.
Week 0 — Stakeholder alignment
Identify a single internal owner for the rollout. Confirm the metric you intend to improve (calls placed per day, hours saved per dispatcher, load-to-driver lead time, settlement cycle time — whichever applies). Align ownership, dispatch leads, and any affected drivers on what's changing and why. This step takes one or two short meetings, not weeks.
Week 1 — Pilot setup
Connect AI Rate Con Upload to a narrow scope first — one dispatcher, one lane, or a subset of customers. Validate that the integration behaves as expected on your real data. Capture any edge cases your operations have that the standard configuration didn't anticipate. X-TMS support is available throughout this phase.
Weeks 2–4 — Scale up gradually
Expand to more dispatchers, more lanes, or higher volume. Most teams scale to full production within 2–4 weeks of the initial pilot. Track the metric you committed to in Week 0; it's the honest signal of whether the deployment is doing what you bought it for.
Ongoing — Iterate
Review AI Rate Con Upload performance monthly with your team for the first quarter. Tune configuration (criteria, thresholds, weights) based on what's working and what isn't. This is normal — every AI-driven workflow benefits from a few iterations as it learns your specific operation.
Common pitfalls to avoid
Skipping the pilot. Teams that try to flip the switch globally on day one tend to discover edge cases at the least convenient moment — under live operational load. A one-week pilot prevents this.
No defined success metric. If you can't articulate what "good" looks like, you won't know whether the deployment succeeded. Pick one number and track it.
Treating AI as a black box. AI Rate Con Upload provides reasoning behind every recommendation. Take advantage of it during the first few weeks — your team learns the AI's logic, and the AI learns your team's preferences.
Frequently asked questions
How accurate is the parsing?
On clean PDFs the engine extracts 90–95% of fields correctly on the first pass. Scanned or low-quality documents drop into the 80–90% range, with low-confidence fields flagged for review.
Can I bulk-upload historical rate cons during migration?
Yes. Customers commonly drop several hundred PDFs at once during onboarding to populate the historical-rate database used for benchmark validation.
Does this work with non-standard rate-con templates?
Yes. The AI doesn't rely on a fixed template — it learns from the document content. Custom or unusual templates work after a brief tuning period.
What languages are supported?
English, Spanish, German, French, Polish, Russian as of June 2026. Other languages are added on request.
How is this different from TED?
TED watches a dedicated inbox for rate cons sent by email. The bulk-upload tool is for documents you already have on disk — both use the same underlying parsing engine.
What happens if the same rate-con is uploaded twice?
The duplicate detector flags it before creating a second load. It compares load number, broker reference, origin, destination, pickup date, and equipment type. If it's a true duplicate, the upload is skipped with a clear notice; if it's a near-match (e.g., a rate amendment), the existing load is updated rather than duplicated.
Can drivers upload BOLs or rate confirmations from the mobile app?
Yes. The X-TMS driver mobile app uses the same parsing engine — drivers can take a photo of a rate-con or BOL on the road and the load record is created or updated immediately. Useful for paper-heavy operations or when brokers hand drivers a physical document at pickup.