Skip to content
ArborBuddy ArborBuddy Help

Import memory & re-process

ArborBuddy keeps import memory in two places: a live hint while you paste leads, and an archived import history you can re-process without uploading again.

Import memory on Lead intake

Section titled “Import memory on Lead intake”

When you type or paste into Lead intake (Yard home or Dispatch → Lead intake), ArborBuddy checks whether the text looks like a client you’ve seen before.

If there’s a strong match, a blue banner appears:

LabelMeaning
Import memory matchToken overlap with a prior CSV row or Accounts directory entry
Vectorize matchSemantic similarity to a prior import (same client phrasing, nearby address)

The banner shows match confidence (%), client name, and property when known. Click View account history to open their Accounts & Sites page.

What triggers a match

Section titled “What triggers a match”
  • Customer or company name in your paste
  • Street address or city
  • Combinations that overlap with import rows or directory records

Vector matches can surface even when wording differs slightly (e.g. “Johnson” vs “Johnson Family — maple removal”). Matches debounce as you type — expect a brief “Checking import history…” state.

What it does not do

Section titled “What it does not do”

Import memory is a hint, not auto-fill. You still review and submit the intake draft. It does not change job data on the board without your approval.

Archived imports & re-process

Section titled “Archived imports & re-process”

Every CSV and batch import is stored in R2 archive tied to your company. The Recent imports panel at the bottom of the Import wizard lists:

  • Filename and timestamp
  • Row count and parse mode (exact, AI mapping, large batch, etc.)
  • Entity summary — clients, sites, jobs created
  • Status — complete, failed, or in progress

Re-process an archived file

Section titled “Re-process an archived file”
  1. Open Import in the sidebar.
  2. Scroll to Recent imports.
  3. Click Re-process on a CSV or batch entry (bulk Magic Paste runs are not re-processable from archive).

ArborBuddy fetches the archived CSV from storage and runs it through the current mapping pipeline again. Use this when:

  • You improved crew setup and want jobs re-staged
  • Column mapping logic was updated since your first import
  • A partial import failed and you want a clean re-run

You do not need to find the original file on your computer.

After re-process

Section titled “After re-process”
  • Preview and import flow is the same as a fresh upload.
  • New import batch record is created; prior board jobs are not automatically deleted — manage duplicates on the dispatch board if needed.

Vectorize & privacy

Section titled “Vectorize & privacy”

Semantic matching uses company-scoped embeddings (Vectorize). Only your workspace’s import and directory data participates in matching. Raw audio from voice intake is not stored — only text transcripts feed intake memory.

Section titled “Related”