Docs
Everything you need to go from zero to alerting in about 5 minutes.
Getting started
- 1
Connect Pipedrive or HubSpot
From the dashboard, click Connect CRM and approve the OAuth flow. We only request read access to deals, pipelines, and owners — we never modify your CRM data. Tokens are stored encrypted at rest (AES-256-GCM).
For Pipedrive, we automatically register a webhook secured with HTTP Basic Auth. For HubSpot, our marketplace app handles delivery — no extra setup.
- 2
Add a delivery channel
Go to Channels → Add channel, choose Google Chat, Slack, Microsoft Teams, or email, and paste the webhook URL (or SMTP config for email). Then click Send test message — you'll need a successful test before you can bind a rule to this channel.
- 3
Create your first rule
Rules → New rule. Pick a trigger (e.g.
deal.won), optionally add conditions ("only for deals over $50k"), and select the channels to fire on. Save and you're live.Start with a template — we ship 10 common setups you can customize.
- 4
Watch the notification log
Notifications shows every alert we processed — delivered, failed, suppressed (quota), or deduplicated. Use it to verify rules are firing and to debug anything that isn't.
Trigger reference
Each rule has one trigger. Conditions narrow which events match; delivery channels control where the alert goes.
| Trigger | Plan | Description |
|---|---|---|
deal.won | All plans | A deal transitions into a closed-won stage. Pipedrive fires this automatically; HubSpot infers it from the pipeline stage metadata. |
deal.lost | All plans | A deal transitions into a closed-lost stage. |
deal.stage_change | All plans | A deal moves to a different stage. Fires on forward and backward movement (also matches deal.stage_regression). |
deal.stage_regression | All plans | A deal moves to an earlier stage in its pipeline. Useful for catching deals that slip backwards. |
deal.stalled | All plans | A deal has been in its current stage longer than your warning/critical/urgent thresholds. Scanned hourly. |
deal.overdue | All plans | A deal's expected close date has passed but the deal is still open. Scanned daily at 9am UTC. |
deal.activity_gap | Starter and up | A deal has had no logged activity in the last N days. Scanned daily at 10am UTC. |
deal.value_change | Starter and up | A deal's value has changed. Useful for high-value deals that get renegotiated. |
digest.weekly_pipeline | Starter and up | Weekly summary of pipeline activity — won, lost, new deals, total value, stalled count. |
digest.rep_performance | Pro and up | Weekly per-rep breakdown of deals closed and pipeline value. |
FAQ
Do I need a HubSpot paid plan for this to work?
No. Nyphex Alerts works with any HubSpot plan — including the free CRM. HubSpot's native workflow-to-Slack alerts require Sales Hub Professional ($100+/seat/month). We give you the same capability starting at $0.
What's actually in the Free plan?
3 alert rules, 1 delivery channel (Google Chat only), 100 notifications per month, 7 days of notification history, 1 CRM connection. No credit card required. Perfect for evaluating or small teams.
How does deduplication work?
We use a 5-minute Redis dedup window per (tenant, deal, event type). When a terminal event like deal.won fires, we cascade-suppress follow-up events (stage change, value change) for the same deal so you get one alert per business change, not five.
What happens if I hit my monthly notification cap?
Further notifications are logged as suppressed (visible in your notification log) but not delivered. Your rules keep running — they resume delivery automatically at the start of the next calendar month. Upgrade to lift the cap.
Is my CRM data stored?
We cache deal metadata (stage, value, owner, activity dates) so scanners like deal.stalled can run without hammering your CRM. We don't store contact or company records. OAuth tokens are encrypted at rest with AES-256-GCM. On disconnect or tenant deletion, all data is cascade-deleted.
Can I trust the alerts to be timely?
Real-time triggers (won/lost/stage_change) deliver in under 5 seconds p95. Scheduled scanners (stalled/overdue/activity_gap) run hourly or daily depending on type. Digests run on your configured schedule. See /status for live health.
How do I get support?
Email support@nyphex.com. Pro and Team plans get priority response.