Asking IQ for a job's file
You do not have to open a job, or even remember which one it was, to get at its record. Ask IQ, and it pulls the job's file into the chat: the same record the [Job file export](/help/jobs-proof-pac…
On this page
The job file, as a conversation
You do not have to open a job, or even remember which one it was, to get at its record. Ask IQ, and it pulls the job's file into the chat: the same record the Job file export compiles into a PDF, read back to you as answers instead of pages. IQ is an AI that works exclusively for you, and on this question it works only from your own record; nothing it reads here involves peers, benchmarks, or anyone else's data.
Two asks that show the shape of it:
- "Pull the file on the Hendersons' water job."
- "Did the adjuster reverse himself on that claim?"
The first is a retrieval: IQ finds the job, confirms it has the right one, and lays out the record. The second is the kind of question the record exists to answer: IQ reads the dispute history and tells you whether a position was reversed, and on what.
What IQ pulls
When IQ opens a job's file, it gets four things:
- The claim identification. Claim number, insured, carrier, client, type of work, status, the money (estimated, approved, billed, collected), and the key dates from assignment through payment.
- The carrier dispute record. For each supplement the carrier pushed back on: which lines were cut, the reasons in the adjuster's own words where they were captured, who the adjuster was and their role, what was submitted against what was approved, the documentation demanded, negotiation rounds and reassignments, whether it was paid on actual cash value, and whether escalation reversed a position. This is the material behind "did the adjuster reverse himself."
- A correspondence summary. The emails linked to the job, each with sender, date, party (adjuster, carrier, customer, sub, or you), subject, and a short excerpt of the body.
- The exhibits list. The documents and attachments on file, numbered as they would appear in the exported index.
If the job has gaps, IQ says so rather than papering over them: no linked emails means the correspondence trail is empty, and IQ will suggest connecting your inbox to build it; no recorded pushback means exactly that, no carrier pushback on this job's supplements.
The bounds, stated plainly
A chat answer is a working summary, not the whole archive, and the bounds are fixed:
- Up to 15 emails are summarized per pull, and IQ is told the true total, so it can say "showing 15 of 42" rather than implying it read everything.
- Each email is excerpted to its first 320 characters, not reproduced in full.
- Up to 40 exhibits are listed, again alongside the true total.
The complete email bodies live only in the exported PDF, and IQ knows it: when you want the whole thing, or a document you can hand to counsel, it points you to the Job file button on the job page, where the full record, complete bodies and exhibits included, exports in one click. A redacted copy for broader sharing is available there too; see sharing a job file safely.
Your data, and only yours
IQ can only pull files for jobs on your own book, and the file it pulls carries no peer, cohort, network, or score figures, the same rule the exported PDF enforces. What the carrier said, what you submitted, what was cut, and what the correspondence shows: all of it is your own record, which is exactly what makes it usable in a dispute.
Data sources
Data sources
- 1.Your job's claim identifiers, amounts, and lifecycle dates. Your business.
- 2.Your supplements and the carrier's line-by-line decisions on them. Your business.
- 3.Emails and attachments linked to this job from your connected inbox. Your business.