Overview
Your tools shouldn’t have to ask what changed - they should just know. Webhooks let RecordsKeeper.AI push events to your endpoint the moment something happens: a record is created or updated, a file is ingested, a sync finishes, or an AI insight is generated. No polling, no delays.
What’s new
Immediate events: get notified the instant key actions occur.
Noise control: subscribe only to the event types you care about.
Built-in security: every delivery is HMAC-signed with your secret.
Production-ready delivery: automatic retries, delivery logs, and clear response handling.
Why this matters
Webhooks turn RecordsKeeper.AI into a live “nerve ending” for your stack. Every change can trigger a downstream action - send a message, open a ticket, refresh a dashboard, update a data store - so teams ship faster with fewer cron jobs and less glue code.
Benefits at a glance
Faster response to customer and data changes
Less polling = fewer wasted API calls and costs
Traceable, secure integrations with signing + logs
Easy to automate across sales, support, data, and compliance
Common use cases (ready-to-copy ideas)
Sales ops: when a contract PDF is ingested, post a summary + link to
#dealsin Slack.Support: on AI insight generated, create a Jira ticket pre-filled with context.
Data/BI: after sync completed, kick your ELT job or refresh a dashboard.
Compliance: if sensitive records change, ping your audit service with the signed payload.
Event types available
Subscribe to any mix of events:
record.createdrecord.updatedfile.ingestedsync.completedai.insight.generated
We’ll continue to add more events; you can update your subscriptions anytime.
How to set it up (≈2 minutes)
In RecordsKeeper.AI, go to Integrations → Webhooks.
Click Create Webhook, choose your event types, and enter your endpoint URL.
Copy the signing secret - you’ll use this to verify requests.
Deploy your endpoint to accept POSTs and return 2xx to acknowledge.
Watch deliveries in Webhook Logs; we retry on transient failures.
Pro tip: start with a staging endpoint to inspect payloads before you go live.
Security & delivery details
HMAC signatures: Each request includes a signature header computed with your webhook’s secret. Verify it before processing.
Retries: We automatically retry on non-2xx responses and network hiccups.
Observability: Delivery attempts and response codes appear in Integration → Webhooks → Logs so you can debug quickly.
Sample flow (high-level)
RecordsKeeper.AI triggers
file.ingested.We POST a JSON payload to your endpoint with headers including the HMAC signature.
Your endpoint verifies the signature, processes the event, and returns 200 OK.
Your app posts to Slack / creates a Jira issue / runs an ELT task—your call.
Quick validation checklist
Endpoint reachable over HTTPS
Signature verification in place using your signing secret
Return 2xx on success; log and handle others
Subscribed only to the events you need
Tested with a staging endpoint before production
FAQs
Does this replace polling?
In most cases, yes. Webhooks are the preferred way to react to changes in real time.
What happens if my endpoint is down?
We’ll retry with backoff. You can review failures in delivery logs and replay after you’re back up.
Can I have multiple webhooks?
Yes - create separate webhooks per environment (staging/production) or per use case.
How do I rotate the signing secret?
Create a new secret, update your service to accept both for a short window, then remove the old one.
Next steps
Create your first webhook in Integrations → Webhooks.
Point a small serverless function at the endpoint, verify the HMAC, and ship an automation in one commit.
If you need help designing your first workflow, tell us your use case (tool, event, desired action) and we’ll share a ready-to-adapt recipe.