The accuracy-first conversion platform for roofing, restoration and public adjusting. Real catalog data. Honest output. Files that import clean.
One wrong code in that manual chain doesn't just look bad — Xactimate re-prices the whole file against its own price list on import. A catch-all code can turn a sane estimate into an astronomical one.
24 of 30 line items carried price-list identity (itemId + skClass) that Xactimate re-resolved against its own price list on import. The reprice dialog told the whole story:
The 9500-class is Xactimate's catch-all (“unassigned”). Emitting it — or stale price-list identity — hands the pricing back to Xactimate on import. The fix isn't luck; it's never emitting unassigned classes and locking the PDF price into the file.
Upload a carrier scope PDF. Get back a finished, validated .esx file — tree, codes, prices and all.
A 12-page carrier scope PDF through the pipeline — in minutes, not hours.
All 20 guard checks green — grand total $33,521.29, 60 ITEMs → 30 SUMITEMs, 0 orphans.
Intentional general-quote export: PDF prices locked, nothing left to reprice.
28/28 checked lines: component sums == PDF unit price, exactly.
The full price list, rebuilt from the source export. This is what makes accurate matching possible.
Built from the user-supplied full price list and real Xactimate exports — rebuilt 2026-08-10 from MASTER.xlsx (AUG26). Verified counts, verified provenance. That's why matching holds up on real scopes.
G1–G18, including: XML well-formedness, ITEM→SUMITEM linkage, price-list linkage, grand-total presence, container shape, component sums, and the G18 catalog-divergence gate — plus the G-PII hard gate: no real customer PII ever lands in files (demo data only).
Every conversion ships with a line-level account — “23 linked / 7 locked” style: which lines matched the catalog, which kept PDF prices and were flagged. No black box.
Public adjusters get a clean depreciation workbook alongside the ESX — the breakdown they need without rebuilding it.
A hard-fail guard keeps real customer PII out of generated files — demo data only, verified on every output.
SOC 2 in progress. AI-assisted conversion with deterministic gates on top. Every claim in this deck is measured and documented — not aspirational.
| Capability | ESXPress | CapOut |
|---|---|---|
| Lines preserved from PDF | All 101 of 101 | 46 (merged) |
| Catalog divergence (G18, on real files) | 0 lines | 21 / 46 |
| Per-conversion accuracy report | ✓ Included | ✗ |
| Depreciation XLSX export | ✓ Included | ✗ |
| Batch conversion | ✓ Included | ✗ |
| EagleView native ESX | In development | — |
| Room-by-room nested structure | ✓ Preserved | Flattened to one group |
Divergence measured on CapOut's own export (MIKEL_CAPOUT_MATH.esx) through our G18 import guard, 2026-08-11. We don't run their pipeline — we measured their output.
Scopes in minutes instead of evenings — and files that get paid, because totals hold.
Supplement volume without accuracy risk — every line preserved, divergence zeroed.
Live in Xactimate. Need defensible files and depreciation breakdowns — both included.
Trust and legitimacy come from honest, verifiable output — which is the only kind we ship.
Standard ESX export · email support
Batch export · profit calculator · priority chat
Photo-scope · API + webhooks · white-label · manager
No cap · everything included
Real catalog data. Honest output. Files that import clean on the first try. The reprice catastrophe is a solved problem — we'd like to prove it to you.
Restoration shops, PA firms, SaaS distributors — integrate conversion into your workflow.
Free tier with 5 conversions, no card. Run your own scope through the 18 gates and read the accuracy report.
A defensible data moat in a market where every estimate ends in Xactimate — and every wrong code is a disaster.