---
title: "How to Move Years of AI Work Out of a Vendor Archive"
description: "Export years of AI conversations, identify the records worth keeping, and turn recurring business context into files your agents can use."
canonical: "https://scalewithsearch.com/articles/move-ai-work-out-of-vendor-archive"
date: "2026-08-17"
modified: "2026-09-19"
---
## Site navigation

- [Scale With Search](https://scalewithsearch.com/)
- Real estate
  - Real estate
    - [Real estate](https://scalewithsearch.com/for/real-estate)
- Work
  - Start here
    - [Send your brief](https://scalewithsearch.com/work#send-your-brief)
    - [Prepare your six-question brief](https://scalewithsearch.com/work#prepare-your-six-question-brief)
  - Build
    - [Site build, content library with SEO, signal desk](https://scalewithsearch.com/work)
- For your business
  - Trades and home services
    - [Auto body and collision shops](https://scalewithsearch.com/for/auto-body-and-collision-shops)
    - [Foundation and home repair contractors](https://scalewithsearch.com/for/foundation-and-home-repair)
    - [Garage door and fencing contractors](https://scalewithsearch.com/for/garage-door-and-fencing-contractors)
    - [HVAC contractors](https://scalewithsearch.com/for/hvac-contractors)
    - [Janitorial and commercial cleaning companies](https://scalewithsearch.com/for/janitorial-and-commercial-cleaning)
    - [Locksmiths](https://scalewithsearch.com/for/locksmiths)
    - [Moving companies](https://scalewithsearch.com/for/moving-companies)
    - [Pest control companies](https://scalewithsearch.com/for/pest-control-companies)
    - [Plumbing and electrical contractors](https://scalewithsearch.com/for/plumbing-and-electrical-contractors)
    - [Restoration and water or fire damage companies](https://scalewithsearch.com/for/restoration-and-water-fire-damage)
    - [Roofing companies](https://scalewithsearch.com/for/roofing-companies)
    - [Towing companies](https://scalewithsearch.com/for/towing-companies)
    - [Tree services and landscaping companies](https://scalewithsearch.com/for/tree-services-and-landscaping)
    - [Solar installers](https://scalewithsearch.com/for/solar-installation)
    - [General contractors](https://scalewithsearch.com/for/general-contractors-and-construction)
    - [Paving, concrete, and flooring contractors](https://scalewithsearch.com/for/paving)
  - Practices and professional services
    - [Bookkeeping and tax practices](https://scalewithsearch.com/for/bookkeeping-and-tax-practices)
    - [Dental practices](https://scalewithsearch.com/for/dental-practices)
    - [Family and criminal defense law firms](https://scalewithsearch.com/for/family-and-criminal-defense-law-firms)
    - [Med spas and aesthetics practices](https://scalewithsearch.com/for/med-spas-and-aesthetics)
    - [Personal injury law firms](https://scalewithsearch.com/for/personal-injury-law-firms)
    - [Veterinary clinics](https://scalewithsearch.com/for/veterinary-clinics)
    - [Gyms and fitness studios](https://scalewithsearch.com/for/fitness)
    - [Therapy and outpatient health practices](https://scalewithsearch.com/for/therapy-and-outpatient-health)
    - [Medical billing companies](https://scalewithsearch.com/for/medical-billing)
    - [Insurance agencies](https://scalewithsearch.com/for/insurance-agencies)
    - [Financial advisors](https://scalewithsearch.com/for/financial-advisors)
    - [Property management companies](https://scalewithsearch.com/for/property-management)
    - [Recruiting and staffing agencies](https://scalewithsearch.com/for/recruiting-and-staffing)
    - [Architects and interior designers](https://scalewithsearch.com/for/architects-and-interior-designers)
    - [Logistics and supply chain companies](https://scalewithsearch.com/for/logistics-and-supply-chain)
  - Agencies, MSPs, and manufacturing
    - [IT and managed service providers](https://scalewithsearch.com/for/it-and-managed-service-providers)
    - [Machine shops and precision manufacturers](https://scalewithsearch.com/for/machine-shops-and-precision-manufacturing)
    - [Marketing agencies and freelancers](https://scalewithsearch.com/for/marketing-agencies-and-freelancers)
    - [SEO agencies and consultants](https://scalewithsearch.com/for/seo-agencies-and-consultants)
    - [Small manufacturers and fabricators](https://scalewithsearch.com/for/small-manufacturers-and-fabricators)
  - Restaurants, shops, studios, and nonprofits
    - [Restaurants and hospitality businesses](https://scalewithsearch.com/for/restaurants-and-hospitality)
    - [Retail stores and ecommerce sellers](https://scalewithsearch.com/for/retail-and-ecommerce)
    - [Photographers, event planners, and travel agents](https://scalewithsearch.com/for/photographers)
    - [Churches and nonprofits](https://scalewithsearch.com/for/churches-and-nonprofits)
  - [All industries](https://scalewithsearch.com/for/)
- Learn
  - For your office
    - [Office job guides](https://scalewithsearch.com/guides/)
    - [Browser calculators](https://scalewithsearch.com/tools/)
  - Start here
    - [How it works](https://scalewithsearch.com/how-it-works)
    - [Free Starter Kit](https://scalewithsearch.com/kit/business-memory-starter-kit.zip)
    - [Synthetic specimen](https://scalewithsearch.com/specimen/working-session-specimen.zip)
  - Guides
    - [The Complete Guide to Business Memory for AI Agents](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide)
    - [The Complete Small-Business Guide to AI Agent Governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business)
    - [The Complete Guide to Leaving Vendor AI Memory](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide)
  - Articles by cluster
    - [Business memory](https://scalewithsearch.com/articles/business-memory-for-ai-agents-guide)
    - [Agent governance](https://scalewithsearch.com/articles/ai-agent-governance-guide-small-business)
    - [Migration and ownership](https://scalewithsearch.com/articles/leaving-vendor-ai-memory-guide)
  - For machines
    - [llms.txt](https://scalewithsearch.com/llms.txt)
    - [llms-full.txt](https://scalewithsearch.com/llms-full.txt)
    - [Machine view](https://scalewithsearch.com/?view=machine)
- Company
  - Evidence
    - [Proof](https://scalewithsearch.com/proof)
  - Company
    - [About](https://scalewithsearch.com/about)

# How to Move Years of AI Work Out of a Vendor Archive.

You open a new [agent](/articles/what-is-an-ai-agent-in-a-small-business) to continue a project that has lived in another AI product for two years. The decisions exist somewhere in hundreds of conversations. The agent cannot distinguish a current rule from an abandoned idea. Before useful work begins, you reconstruct the business again.

Downloading the archive does not solve that incident. It gives you custody of exported data. The next job is to turn the useful parts into a [business record](/articles/stop-re-explaining-your-business-to-ai).

## Start with an export, not a copy-and-paste marathon

Use the vendor's documented export process first. [OpenAI documents](https://help.openai.com/en/articles/7260999-how-do-i-export-my-chatgpt-history-and-data) export through Data Controls or its Privacy Portal for eligible accounts. The downloaded ZIP includes chat history and other account data. [Anthropic documents](https://support.claude.com/en/articles/9450526-export-your-claude-data) exports for eligible Claude accounts; check the role and account requirements before requesting a workspace export.

Those are acquisition steps. They do not promise that another product can import every conversation with the same structure, memory, attachments, or behavior.

Preserve the original ZIP unchanged. Record its filename, request date, receipt date, account or workspace, and a checksum if your environment already uses checksums. Do not rewrite the only source copy while you are still deciding what matters.

Create an inspectable file named `archive-inventory.md`. It should identify:

- the vendor and account scope;
- the export request and receipt dates;
- the original archive filename;
- the formats found inside;
- the earliest and latest conversation dates;
- the recurring jobs you expect to recover;
- known missing workspaces, projects, or attachments;
- material that must remain restricted;
- the person who can approve deletion or wider access.

This file is the first receipt. It tells a future reader what arrived and what did not.

## Build a migration register before extracting rules

The inventory describes the archive. A migration register describes the decisions made about its contents. For naming conversations so the original evidence stays findable, read [Chat History Title Best Practices for Business AI.](/articles/chat-history-title-best-practices-business-ai)

Create one row for each recurring job or source group you may promote:

```text
item_id:: MIG-012
archive_location:: conversations/weekly-client-review-2025.html
candidate_job:: prepare weekly client review
data_owner:: account owner
contains_restricted_data:: unknown
current_authority:: unverified
target_packet:: clients/CL-001/weekly-review/
review_status:: pending
migration_status:: not started
```

Do not use a model-generated summary as the only register. The register needs stable archive locations so a reviewer can return to the source. It should show skipped material and unresolved authority, not only successful extraction.

This also creates a stopping point. If ownership, sensitivity, or current authority is unknown, the item remains pending. It does not enter working context because the archive happened to contain it.

## Separate custody from usefulness

An archive contains at least three different kinds of material.

The first is source material: documents, facts, policies, customer records, research, and links that came from outside the model.

The second is operating knowledge: decisions, corrections, preferred formats, approval rules, and stopping points developed during the work.

The third is generated residue: drafts, abandoned approaches, duplicated answers, speculative claims, and conversations that never became a decision.

Do not promote all three into permanent business memory. A large pile of chats is hard to inspect and easy to misread. Extraction requires judgment.

For each recurring job, build a small migration packet:

- `sources.md` lists confirmed source records and their locations;
- `decisions.md` states current decisions and superseded alternatives;
- `corrections.md` records errors that should not recur;
- `task-brief.md` defines the job, required input, output, and acceptance conditions;
- `boundaries.md` names what the agent may do and where it must stop;
- `migration-receipt.md` links each promoted statement back to its archive location.

The archive remains evidence. The smaller packet becomes working context.

## Migrate one recurring job first

Do not begin by converting every conversation. Pick the job that causes the most repeated reconstruction.

Suppose you repeatedly ask an agent to prepare a client review. The old archive contains client facts, a reporting format, several corrections, and discussions about what may be sent. Extract only the current records required for that review.

Write an acceptance test before moving more material:

1. Start a fresh session with no old chat open.
2. Give the agent only the migration packet and current task input.
3. Ask it to produce the defined artifact.
4. Check every factual statement against `sources.md`.
5. Check the format against `task-brief.md`.
6. Confirm that it stops before any external send or record change.
7. Record misses in `corrections.md` and run the test again.

The inspectable result is `migration-test-01.md`. It should contain the input set, model and tool used, output location, acceptance checks, failures, corrections, and final disposition.

Passing this test does not prove the whole archive is portable. It proves one job can run from records you control.

## Reconcile the packet with the archive

After the first test passes, check the promoted packet against its source locations.

Every current decision should be buyer-confirmed or traceable to the archive. Every superseded rule should be marked so it cannot re-enter through later extraction. Every correction should name the behavior it changes. Every excluded source should remain outside the packet.

Write the reconciliation result into `migration-receipt.md`:

```text
job:: weekly client review
archive_items_reviewed:: 7
records_promoted:: 3
records_rejected_as_outdated:: 2
records_pending_owner_decision:: 2
test_result:: pass with pending items excluded
archive_deleted:: no
next_scope_decision:: buyer review required
```

A passed job test does not authorize deleting the archive. The archive may still be needed for provenance, dispute review, or later migration. Deletion has its own owner, retention rule, and approval.

## Migrate decisions, not conversational style

Long-running conversations contain turns of phrase, speculative branches, and model-generated explanations. Most do not belong in the source record. For selecting the durable facts without promoting the whole archive, read [What to Keep From an AI Chat Archive: A Review Test.](/articles/what-to-keep-from-ai-chat-archive)

Promote current facts, accepted decisions, reusable corrections, explicit exclusions, and job boundaries. Keep a small number of approved examples only when format matters. Label examples as structure-only when their facts cannot be reused.

This keeps the migration packet small enough to inspect. The goal is to run the job without the old interface, not to make the new system imitate every old conversation.

## Keep identity and permissions visible

Vendor archives can contain material from several clients, roles, or businesses. Do not combine them into one context folder because they came from one account.

Route each source to the correct owner and access boundary. Personal material stays out of client context. One client's records stay out of another client's task brief. Credentials do not belong in Markdown or migration receipts. Record secret names and custody locations, never secret values.

If the export belongs to a company workspace, confirm who has authority to extract and retain it. An available export button is not a permission grant for every conversation inside the archive.

## Define the stop point before extraction

The migration system must stop and ask for approval when it encounters:

- unclear ownership;
- personal or regulated data outside the stated scope;
- a conflict between two current-looking decisions;
- a request to delete the vendor archive;
- a request to share extracted records with a new person or system;
- an external action such as sending, publishing, booking, or changing a live record.

It should also stop when the first recurring job passes its [transfer](/articles/transfer-the-repo-client-handoff) test. Broader migration is a new scope decision, not an automatic next step.

## What ownership looks like after migration

The finished object is not a cleaned transcript collection. It is a bounded set of readable files with sources, decisions, corrections, approval rules, and a test receipt.

You can still keep the original export for evidence. You can still search old conversations when a question appears. The difference is that the next agent does not need the old vendor interface to start the recurring job.

Custody comes first. Usable memory comes from selection, structure, and testing.

## Sources

- [OpenAI: How to export ChatGPT history and data](https://help.openai.com/en/articles/7260999-how-do-i-export-my-chatgpt-history-and-data)
- [Anthropic: How to export Claude data](https://support.claude.com/en/articles/9450526-export-your-claude-data)


## Questions about How to Move Years of AI Work Out of a Vendor Archive

### How do you move ChatGPT content into Obsidian?

Preserve the vendor export as an unchanged source copy, then create an archive inventory and migration register. The working packet should contain confirmed sources, decisions, corrections, task boundaries, and a migration receipt rather than every conversation.

### How to Transfer your Folders/Projects from GPT to Another LLM?

Start with the vendor's documented export process and record what the archive contains and omits. Separate source material, operating knowledge, and generated residue, then migrate one recurring job into a small packet that another operator can inspect.

### How can I export ~850MB of ChatGPT conversations to migrate to another AI model?

A large export is custody of data, not proof that a job is portable. Freeze one test packet, run the job from a clean session, check the output against current sources, and record the result before expanding the migration.

## Related: Ownership and Migration

- [What to Keep, Reject, or Review From an AI Chat Archive](/articles/what-to-keep-from-ai-chat-archive)
- [Convert a ChatGPT Export to Markdown Without Making Every Chat Truth](/articles/chatgpt-export-to-markdown-business-memory)
- [Replace the Model Without Rebuilding the Business Memory](/articles/replace-the-model-keep-business-memory)

----

```text
                  .|########||.                                       .|########||.                                       .|########||.
               |##||.      .||##|.                                 |##||.      .||##|.                                 |##||.      .||##|.
             |#|.              .|#|.                             |#|.              .|#|.                             |#|.              .|#|.
           |#|                    |#|                          |#|                    |#|                          |#|                    |#|
         .#|                        |#.                      .#|                        |#.                      .#|                        |#.
        .#.                          .#|                    .#.                          .#|                    .#.                          .#|
       |#.                            .#|                  |#.                            .#|                  |#.                            .#|
      |#             ......             #|                |#             ......             #|                |#             ......             #|
     .#           ||#########|           #|              .#           ||#########|           #|              .#           ||#########|           #|
    .#.         |######||######|.        .#.            .#.         |######||######|.        .#.            .#.         |######||######|.        .#.
    #.        .##|###|##|#|######|        .#            #.        .##|###|##|#|######|        .#            #.        .##|###|##|#|######|        .#
   ||        |##|#||||||||||||#||#|        ||          ||        |##|#||||||||||||#||#|        ||          ||        |##|#||||||||||||#||#|        ||
   #        |#||||||||||||||||||||#|        #.         #        |#||||||||||||||||||||#|        #.         #        |#||||||||||||||||||||#|        #.
  ||       |#||||||||||||||||||||||#|       ||        ||       |#||||||||||||||||||||||#|       ||        ||       |#||||||||||||||||||||||#|       ||
  #       .#||||||||||||||||||||||||#|       #        #       .#||||||||||||||||||||||||#|       #        #       .#||||||||||||||||||||||||#|       #
 ||   ....|||#||||||##|#|||#|#||##||||....|. ||      ||   ....|||#||||||##|#|||#|#||##||||....|. ||      ||   ....|||#||||||##|#|||#|#||##||||....|. ||
 #.  .  ....|#  ....#|||   ||| .#|||  ....#. .#      #.  .  ....|#  ....#|||   ||| .#|||  ....#. .#      #.  .  ....|#  ....#|||   ||| .#|||  ....#. .#
 #   .  ||||#| .#####||  . .#| .#|||  ||||#   #.     #   .  ||||#| .#####||  . .#| .#|||  ||||#   #.     #   .  ||||#| .#####||  . .#| .#|||  ||||#   #.
.|   |||||  #. |#|||||. ||  #. |#||. .|||||   ||    .|   |||||  #. |#|||||. ||  #. |#||. .|||||   ||    .|   |||||  #. |#|||||. ||  #. |#||. .|||||   ||
||   |....  #. ....|#.      |. ...|. ....||   ||    ||   |....  #. ....|#.      |. ...|. ....||   ||    ||   |....  #. ....|#.      |. ...|. ....||   ||
#.  .||||||##||||||##||####|||||||#||||||#|   .#    #.  .||||||##||||||##||####|||||||#||||||#|   .#    #.  .||||||##||||||##||####|||||||#||||||#|   .#
#    .||####|###########################|.     #    #    .||####|###########################|.     #    #    .||####|###########################|.     #
#      .#||||||.#.|| # |. ..# #| #|||||#|      #    #      .#||||||.#.|| # |. ..# #| #|||||#|      #    #      .#||||||.#.|| # |. ..# #| #|||||#|      #
#      .#|||||| ..  |# ## |#| ...#|#|||#|      #    #      .#|||||| ..  |# ## |#| ...#|#|||#|      #    #      .#|||||| ..  |# ## |#| ...#|#|||#|      #
#      .#|||#|# .# .#| #| ##|.#..#|||||#|      #    #      .#|||#|# .# .#| #| ##|.#..#|||||#|      #    #      .#|||#|# .# .#| #| ##|.#..#|||||#|      #
#      .#||||||############|######|||||#|      #    #      .#||||||############|######|||||#|      #    #      .#||||||############|######|||||#|      #
#   |...||#||||||#||||#|#||||||#|||||#||| |.|  #    #   |...||#||||||#||||#|#||||||#|||||#||| |.|  #    #   |...||#||||||#||||#|#||||||#|||||#||| |.|  #
#. .. ||||# .|||##|.  |#| .|| || .|||#. #|  # .#    #. .. ||||# .|||##|.  |#| .|| || .|||#. #|  # .#    #. .. ||||# .|||##|.  |#| .|| || .|||#. #|  # .#
|| |  ...||  ...##| |  #| .|. |. #####  .. .| ||    || |  ...||  ...##| |  #| .|. |. #####  .. .| ||    || |  ...||  ...##| |  #| .|. |. #####  .. .| ||
|| ||||| || ||||#|  .  |. |. |#. ||||| |#| || ||    || ||||| || ||||#|  .  |. |. |#. ||||| |#| || ||    || ||||| || ||||#|  .  |. |. |#. ||||| |#| || ||
.# |.....#|....|#.||||.|||##.|#|....||.#||.#. #.    .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #.    .# |.....#|....|#.||||.|||##.|#|....||.#||.#. #.
 #.|||||||#############################| ||| .#      #.|||||||#############################| ||| .#      #.|||||||#############################| ||| .#
 ||       ##||#||#||||||#||#||#||#||##.      ||      ||       ##||#||#||||||#||#||#||#||##.      ||      ||       ##||#||#||||||#||#||#||#||##.      ||
  #       .#|||||||||#||#|||||||||||#|       #        #       .#|||||||||#||#|||||||||||#|       #        #       .#|||||||||#||#|||||||||||#|       #
  ||       |#||||||||||||||||||||||#|       ||        ||       |#||||||||||||||||||||||#|       ||        ||       |#||||||||||||||||||||||#|       ||
  .#        |#||||||||||||||||||||#|        #.        .#        |#||||||||||||||||||||#|        #.        .#        |#||||||||||||||||||||#|        #.
   ||        |#||||||||||||||||||#|        ||          ||        |#||||||||||||||||||#|        ||          ||        |#||||||||||||||||||#|        ||
    #.        .######|#|#########|        .#            #.        .######|#|#########|        .#            #.        .######|#|#########|        .#
    .#.         |#####||||#####|         .#.            .#.         |#####||||#####|         .#.            .#.         |#####||||#####|         .#.
     |#           ||########||           #.              |#           ||########||           #.              |#           ||########||           #.
      |#             ......             #|                |#             ......             #|                |#             ......             #|
       |#.                            .#|                  |#.                            .#|                  |#.                            .#|
        |#.                          .#.                    |#.                          .#.                    |#.                          .#.
         .#|                        |#.                      .#|                        |#.                      .#|                        |#.
           |#|                    |#|                          |#|                    |#|                          |#|                    |#|
            .|#|.              .|#|.                            .|#|.              .|#|.                            .|#|.              .|#|.
               |##||.      .||##|                                  |##||.      .||##|                                  |##||.      .||##|
                 .||########||.                                      .||########||.                                      .||########||.

Scale With Search  2026  [scalewithsearch.com](https://scalewithsearch.com)
```