Project Management

Harvest Data Extractor

Vern's Harvest extractor imports a new user's project management data — invoices, contacts, users, and projects — into your product through a single managed migration, handling Harvest's auth, pagination, and rate limits for you. It's built for teams onboarding customers off Harvest who want their data in your app from day one.

Start migration

Browse all extractors

Harvest

SaaS API (REST)

Managed auth, retries, and throttling

Normalized JSON, CSV, and DB-ready output

Cleaned JSON, CSV, and DB-ready output

Export coverage

Know exactly what Vern pulls before you migrate.

Objects Vern extracts

Vern extracts the full set of queryable Harvest objects (32 total), including: Billable Rates, Clients, Company, Contacts, Cost Rates, Estimate Item Categories, Estimate Messages, Estimates, Expense Categories, Expenses, Expenses Categories, Expenses Clients, Expenses Projects, Expenses Team, Invoice Item Categories, Invoice Messages, Invoice Payments, Invoices, Project Assignments, Project Budget, Projects, Roles, Task Assignments, Tasks, Time Clients, Time Entries, Time Projects, Time Tasks, Time Team, Uninvoiced, User Assignments, and Users. 28 of these support incremental change tracking, so syncs pull only what changed.

Output schema

Normalized JSON per object with referential keys preserved, so the relationships between your user's records survive the import. Timestamps are returned in ISO 8601. Delivered as JSON or CSV, or loaded straight into your own application database or pushed to your API — whatever your product expects.

Where it can land

Your own application database (Postgres, MySQL), a direct push to your API or webhook, or JSON/CSV files, plus another project management system (Aha, Basecamp, ClickUp) when a user is switching tools.

Source details

Built for the source type shown above, with extractor content refreshed from the collection record.

Managed by Vern

The details you normally have to build yourself.

Authentication

OAuth 2.0 (Authorization Code flow). Your new user authorizes Harvest once during onboarding — their credentials never touch your code. Harvest access tokens are short-lived and refresh tokens roll on the provider's own schedule, so Vern manages the full token lifecycle and refresh and the import never drops midway.

Pagination

Vern detects and handles Harvest's paging model (cursor, page-number, or offset) internally and returns one consistent, de-duplicated result set — so you import a user's complete data without ever managing page tokens.

Rate limits and gotchas

Harvest enforces its own API rate limits. Vern queues and throttles requests internally so importing a new user's data never trips a 429 — even on a large account.

OAuth-based sources expire access tokens frequently and require a working refresh flow — a dropped refresh is the most common cause of a half-finished import. Vern owns the refresh loop so a new user's onboarding completes even on a long extraction.

API quickstart

1. Create a migration for Harvest
curl -X POST https://app.vern.so/api/v1/migrations \
  -H "x-api-key: $VERN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "name": "Harvest → Postgres", "source": "Harvest" }'
→ { "migration_id": "mig_..." }
2. Generate the extraction recipe (pulls live via the Harvest API)
curl -X POST https://app.vern.so/api/v1/migrations/$MIGRATION_ID/runs \
  -H "x-api-key: $VERN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "kind": "generate", "extract_via_api": true }'
3. Run the approved extraction
curl -X POST https://app.vern.so/api/v1/migrations/$MIGRATION_ID/runs \
  -H "x-api-key: $VERN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{ "kind": "execute" }'
4. Download the normalized CSV export
curl https://app.vern.so/api/v1/migrations/$MIGRATION_ID/exports/{template} \
  -H "x-api-key: $VERN_API_KEY" -o harvest_export.csv

A cleaner way to extract without rebuilding source-specific plumbing.

View docs

Common questions

Can I keep a user's Harvest data in sync after the first import? Yes — 28 Harvest objects support incremental change tracking, so after the first full import Vern pulls only what changed on re-run.

Can I move a user from Harvest to Aha? Yes — Vern normalizes the Harvest data model into typed tables that map cleanly onto Aha and into your own product.

What does it cost to read the Harvest API? Harvest's own API access follows Harvest's pricing/plan; Vern is billed separately per import. Check Harvest's developer docs for their current limits.