Install the agent, ship your first logs and build a live dashboard before your coffee cools.
You do not need a migration plan to try Helix. You need one endpoint and one event. Here is the whole thing.
Send an event
Any structured JSON over HTTPS works. Tag it with a service name so it groups cleanly.
curl -X POST https://ingest.helix.dev/v1/logs \
-H "authorization: Bearer $HELIX_KEY" \
-d '{"service":"checkout","level":"info","msg":"order placed"}'
Query it back
Open HelixQL and filter to your service. Your event is already there.
Wire it into CI
Once a single event lands, the same endpoint drops into your deploy pipeline: emit a structured deploy event on every release and you get a free annotation layer across every dashboard. Now "what changed at 14:02?" answers itself instead of starting a thread.
From five minutes to a habit
The CLI is the fast path in, but the point is the loop: instrument, query, learn, repeat. Keep the queries you run during an incident and the next responder starts where you finished.