Managed service feel. Your infrastructure. Here's exactly what that means.
Architecture
Landed is software that runs inside your cloud account. It's not a SaaS platform that your data passes through — it's a service you license that runs on your own compute, writes to your own warehouse, and never phones home with your data.
Data flows directly from source to your warehouse. Landed never sees your data — it only orchestrates the connection. Connector updates are pulled from our servers, but your data never leaves your VPC.
Getting started
Landed ships as a Docker container. You run it in your own GCP or AWS account — one command and it's up. It needs outbound internet access to reach source APIs, and write access to your destination warehouse. That's it.
# Pull and start Landed docker-compose up -d landed # Verify it's running landed status
Add connectors through a simple YAML config or the CLI. For OAuth-based sources, Landed opens a browser flow — you authenticate directly with the source, and the token is stored locally. We never see it.
# connectors.yaml connectors: - name: bing-ads type: bing_ads account_id: "12345678" destination: bigquery schedule: "*/5 * * * *" # every 5 min
On each sync, Landed pulls from the source API, validates the data, and writes it to your destination. Every sync produces a receipt — row counts, schema checks, and a timestamp. You always know exactly what landed and when.
_landed_at, _is_deleted, and _is_current to every rowSource APIs change. Fields get renamed, endpoints get deprecated, auth flows get updated. Landed checks for connector updates on a regular cadence and applies them automatically — the same way your OS applies security patches. You don't have to do anything.
Principles
This is non-negotiable. Landed is software you run, not a platform your data passes through. We have no access to your data, your credentials, or your customers' information.
A scheduled API call costs pennies to run. We charge a flat rate per connector — not a percentage of your data volume, not a per-row fee, not a MAR limit. You pay for the software, not the compute.
You're running on your own compute. If you want data every minute, sync every minute. We don't throttle frequency or charge more for it. The only limit is what your source API allows.
When an upstream API changes, your connector should update before it breaks — not after you file a support ticket. We monitor API changelogs and ship fixes proactively.
Every sync produces a receipt. Row counts, schema checks, primary key validation. If something's wrong, you find out immediately — not when someone asks a question your data can't answer.
Getting started is easy. But we also care about day 90 — monitoring, debugging, schema migrations, adding new connectors. The dashboard and CLI are designed for ongoing operation, not just setup.
Common questions
Join the early access list. We'll reach out personally and walk you through setup.