PostgreSQL reporting,without waiting on an engineer.

Ask the Connector Agent on the Connect page for PostgreSQL, or select it in the catalog. Enter the host, database, username and password in the secure form (the port defaults to 5432 and SSL is on), pick a schema and tables, then approve the load and a schedule.

You choose one schema per connection, then take every table in it or select specific ones. Base tables, views, materialized views and foreign tables are all listed, and you can add your own SQL queries as extra sources. Querri only sees what the connected Postgres user is allowed to read.

Full refreshes run on a daily, weekly or monthly schedule. An incremental schedule can run as often as hourly and pulls only rows whose cursor column, such as updated_at, is newer than the last run. A sync you start by hand is a full sync. Incremental runs don't see deleted rows; a full refresh does.

No. Querri only reads your data. It attaches the database in read-only mode, so it can't insert, update or delete a row or alter a table. For extra assurance, connect with a Postgres user that only has SELECT rights.

Questions never run against your database; they run on the copy Querri syncs. The sync itself is tuned to go easy, with a handful of connections per table, and you choose when it runs. You can also point it at a read replica so the primary isn't touched at all.

Yes. Enter the replica's host instead of the primary's. Keep in mind that a hot standby cancels queries that conflict with replication after max_standby_streaming_delay, 30 seconds by default, so a sync of a large table may need that limit raised. Turning on hot_standby_feedback prevents the cancellations caused by vacuum cleanup, at the cost of some bloat on the primary.

Querri can only sync the schemas and tables the connected Postgres user can read, and the schema picker only shows schemas that user has USAGE on. Once data is in Querri, row-level access and sharing settings in Querri decide what each person sees; Postgres roles don't carry over per person.

Use the VPN tunnel panel in the connection form and upload your own OpenVPN profile, CA certificate, client certificate and private key. All four are stored encrypted. SSH tunnels aren't supported for PostgreSQL today, so the VPN is the route for databases that aren't reachable from the internet.

Connect both, and Querri joins them in the Library on a shared key such as an account ID or email domain. You can then put product usage next to renewal dates, account owners and pipeline, and when a match is uncertain the answer says so instead of guessing quietly.

Connect Postgres and HubSpot, and Querri matches app workspaces or users to HubSpot contacts and companies. That lets you see which active trials have no deal, which leads ever activated, and which marketing sources bring users who stick around.

Connect both, and Querri lines up support conversations with what each account does in your product. Ask which features drive the most conversations, or whether accounts that contact support early activate faster or slower than the rest.

Tell Querri which event counts as first value, and it measures the time from each account's signup to that event, by week, plan or channel. The signup-to-first-value playbook walks through the steps, and the result can run as a KPI that keeps its history.

Teach Querri the rules once: which status means canceled, which rows are soft-deleted, which accounts are internal. Every count after that follows them. When churn could mean cancellations or ended plans, Querri says which reading it used and shows the other.

Build the analysis once as a project in Querri, then schedule it. Each run uses the latest synced data and sends a formatted Excel workbook by email, or one version per team or customer, so nobody runs a query and pastes results by hand.

Yes. Describe the dashboard you want, or start from a screenshot of one you already have, and change it by asking. It can mix Postgres with billing, CRM and support data, and each person or customer can be limited to their own rows.

Yes. Link Querri with Claude or ChatGPT one time. The assistant answers from your Querri Library with your team's definitions, under each person's own access, and never gets a database login. It works from the same definitions your dashboards use.

Each connection reads one schema, so several schemas mean several connections. Custom SQL can't sync incrementally, incremental runs only on a schedule, and it doesn't detect deleted rows. There's no SSH tunnel, no SSL certificate verification beyond require, and no change-data-capture streaming; data arrives on a schedule.

Yes. If the user can't read Postgres's system catalogs, Querri falls back to a simpler listing. In that mode materialized views aren't listed and custom SQL queries are refused, so select the tables and views you need directly.

Querri is SOC 2 Type II, HIPAA and ISO 27001:2022 certified, and your data is never used to train a model. Connections require SSL by default, passwords and VPN files are stored encrypted, the Connector Agent never sees them, and the database is attached read-only.