ConvOps ConvOps
Docs Alerts Guide Log in →
ConvOps ConvOps

Stop debugging alerts for 45 minutes. Fix them in 2.

ConvOps investigates your AWS alerts, finds the root cause, and lets you fix issues instantly — even at 2am, from your phone.

Get Started Free

Here's exactly what happens during an incident

Real alert → AI diagnosis → root cause + suggested fix. No laptop needed.

1 🔍 Diagnosis
2 ⚡ Action
3 ✅ Resolution
🤖
ConvOps Agent
🔍 DIAGNOSIS
ConvOps Infrastructure Alert
ConvOps Alert:

Resource: api-service
Metric: ALARM - Threshold: > 90%
Env: production | Region: eu-central-1

Analysis: CPU 96%, 3× traffic spike after deploy 12 min ago. No OOM errors. Likely traffic overload, not a memory leak.
Impact: Service degraded. Response times rising. Risk of task failure if load continues.
Causes: Deploy v2.4.1 triggered traffic spike. Tasks under-provisioned for current load.
Suggested: Scale out — add 2 tasks to absorb traffic spike.
Recommendation: ECS adds 2 tasks (~60s). CPU expected to drop from 96% → ~54%.

Reply:
1 - Take Action
2 - Snooze
3 - Ignore
4 - Investigate
5 - Escalate

Alert ID: ALT-20260325-0214 | www.convops.io
02:14
⚡ ACTION
1
02:15
⚠️ Confirm: scale api-service from 2 → 4 tasks on prod-cluster?

Reply YES to confirm or NO to cancel. (expires in 5 min)
02:15
YES
02:15
✅ RESOLUTION
✅ Scaling out. 2 new tasks deploying now.

CPU: 96% → 71% → 54%
All 4 tasks healthy. Action logged.
02:17
# infra-alerts
🔍 DIAGNOSIS
🤖
ConvOps2:14 AM
ConvOps Infrastructure Alert
ConvOps Alert:

Resource: api-service
Metric: ALARM - Threshold: > 90%
Env: production | Region: eu-central-1

Analysis: CPU 96%, 3× traffic spike after deploy 12 min ago. No OOM errors. Likely traffic overload, not a memory leak.
Impact: Service degraded. Response times rising. Risk of task failure if load continues.
Causes: Deploy v2.4.1 triggered traffic spike. Tasks under-provisioned for current load.
Suggested: Scale out — add 2 tasks to absorb traffic spike.
Recommendation: ECS adds 2 tasks (~60s). CPU expected to drop from 96% → ~54%.

Reply: 1 - Take Action · 2 - Snooze · 3 - Ignore · 4 - Investigate · 5 - Escalate

Alert ID: ALT-20260325-0214 | www.convops.io

⚡ ACTION
N
Alex2:15 AM
1
🤖
ConvOps2:15 AM
⚠️ Confirm: scale api-service from 2 → 4 tasks on prod-cluster?

✅ YES — Scale out ❌ NO — Cancel
N
Alex2:15 AM
YES

✅ RESOLUTION
🤖
ConvOps2:17 AM
✅ Scaling out. 2 new tasks deploying now.

CPU: 96% → 71% → 54%
All 4 tasks healthy. Action logged.

That's ConvOps. From alert to resolved — without opening a laptop.

The Problem

3 AM. CloudWatch fires. You get: "CPUUtilization > 80%"

That's it. Now you open your laptop, SSH in, grep logs, check 4 dashboards, and try to figure out what actually broke — half asleep.

ConvOps does that investigation before it wakes you. You get root cause, not a raw metric.

Without ConvOps vs With ConvOps

Without ConvOps
  1. Open CloudWatch console
  2. Check metrics manually
  3. Search through logs
  4. Guess root cause
  5. Execute fix manually — hope it works
⏱ 30–60 minutes. Laptop required.
With ConvOps
  1. Get alert with diagnosis already done
  2. See root cause instantly
  3. Reply 1 to fix — done in seconds
⚡ Under 2 minutes. No laptop.

Up and Running in 2 Minutes

No agents to install. No SDK. No DevOps needed to set up DevOps.

1

Sign up & connect your AWS account

~2 min

Create your account at app.convops.io. Enter your AWS Account ID and region. Click Launch Stack — ConvOps deploys one CloudFormation template that creates an IAM role and SNS topic. No CLI, no credentials, no agents.

Read-only access by default — no changes to your infra without your approval
2

Choose where to receive alerts

Your choice

Pick WhatsApp, Slack, or both — wherever your team already is. Enter your number or Slack webhook. No new tools to learn. Alerts land in the app you already have open.

💬 WhatsApp # Slack
3

CloudWatch fires — AI investigates instantly

Auto

When an alarm triggers, ConvOps AI reads your CloudWatch logs, metrics, and recent events before paging you. You get a message with root cause analysis already done — not just "CPU is high."

4

Reply to act — you stay in control

Human-in-loop

Reply YES to restart a service, reboot an instance, or scale out. Reply NO to cancel. Every action is logged with who approved it, when, and what changed. No autonomous execution — ever.

No action is taken without your explicit approval — confirmation required every time
Request Early Access — Free →

14-day free trial. No credit card required.

How it works during a real incident

ECS service error spike — 2:13 AM

🚨
Alert received
CloudWatch fires: api-service · ErrorRate > 5% · 3 min
🔍
ConvOps investigates automatically
  • Pulls CloudWatch logs → finds OutOfMemoryError in container logs
  • Correlates with memory metric spike 4 min before errors started
  • Identifies failing container: api-worker-3
  • Checks other tasks — 3 of 4 healthy, 1 in STOPPED state
💬
You receive on WhatsApp / Slack
"🔴 api-service error spike. Root cause: memory leak in api-worker-3 (OOM). 1 task stopped. Suggested fix: restart service to cycle the container. Reply 1 to restart, 2 to monitor only."
You reply 1 — service restarts
ConvOps confirms, executes, reports back: all 4 tasks healthy. Total time: 2 minutes. Audit log written.
💤
Back to sleep
No laptop. No VPN. No 5 browser tabs. Laptop stayed closed.

Try it yourself — no account needed

Click simulate to see exactly what ConvOps sends you during a real incident.

🤖
ConvOps

ECS error spike or Lambda timeout — randomised each time

Your infrastructure. Your control.

ConvOps has access to your AWS account. Here's exactly what that means — no vague claims.

Read-only by default

ConvOps reads CloudWatch metrics, logs, and resource state. No write access until you explicitly enable it. You choose which services can be acted on.

Manual approval required

No action executes automatically. Ever. You reply YES to confirm, NO to cancel. Confirmation is phone-bound — only you can approve your alerts. Expires after 5 minutes.

Full audit logs

Every investigation and action logged with timestamp, resource ARN, who approved it, and outcome. 90-day retention. You can export anytime.

Scoped IAM permissions

One CloudFormation template. Minimum permissions needed — nothing more. Review or delete the IAM role anytime from your AWS console.

15-min STS sessions

Tokens expire after 15 minutes. No long-lived credentials stored outside your AWS account.

Revoke instantly

Delete the IAM role from your AWS account. ConvOps loses all access immediately — no support ticket needed.

“If you're an engineer handling production
— getting paged for alerts that tell you nothing until you've
spent 45 minutes in the console — ConvOps was built for you.
No dedicated SRE. No PagerDuty budget. Just you, your phone, and AWS.”

What's free vs paid

See what's wrong for free. Fix it instantly with Pro.

🔍 Free Always free
  • Alert analysis
  • Root cause insights
  • ~ 50 incidents/month
Pro $49/mo
  • Execute fixes from chat
  • Automation rules
  • Team workflows
Get Started Free

Pricing

Start free. Alerts + full AI investigation included. No card required.

Pro & Enterprise with action execution coming soon. Free plan available now.

Pro costs less than 1 hour of engineer time per month.

Free

$0/mo

No credit card required

  • 50 incidents/month
  • 10 investigations/month
  • 10 actions/month
  • Connect 1 AWS account
  • CloudWatch alerts via WhatsApp or Slack
  • Full AI triage
Get Started Free

Enterprise

Custom

For companies with compliance needs

  • Everything in Free
  • ⏳ Everything in Pro
  • ⏳ RCA export to PDF/Jira/Confluence
  • ⏳ Audit logs + SSO + SCPs
  • ⏳ SLA + dedicated support
  • ⏳ Custom integrations
  • ⏳ GDPR DPA included
Contact Us

Free now — alerts & AI investigation. Pro & Enterprise with action execution coming soon.

FAQ

Is this safe to connect to my AWS account? ConvOps deploys a single CloudFormation template that creates:
  • An IAM role with AWS-managed ReadOnlyAccess policy (read-only across all services)
  • An SNS topic to receive CloudWatch alarm notifications
  • An EventBridge rule to route alarm state changes
It cannot modify, delete, or create any AWS resource in the free plan. The IAM role is protected by a unique ExternalId (prevents unauthorized access) and has a 1-hour max session duration. Action execution (Pro) requires explicit human approval for every action — ConvOps never runs anything autonomously. You can revoke access instantly by deleting the IAM role.
How long does the AI investigation take? Under 60 seconds from when CloudWatch fires the alarm to when you receive the RCA on WhatsApp or Slack.
What happens when I hit the 50 incident limit on the free plan? You stop receiving AI investigations for the rest of that month. Your CloudWatch alarms still fire normally — ConvOps just won't enrich them until the next billing cycle. You'll get a warning message when you're approaching the limit.
What if ConvOps has downtime during an incident? Your CloudWatch alarms fire independently of ConvOps. If ConvOps is unavailable, you receive a standard CloudWatch notification via SNS as a fallback. Your monitoring doesn't depend on ConvOps being up.
Is this GDPR compliant? Where is my data stored? ConvOps processes CloudWatch log data in the same AWS region as your account. Log content is not stored beyond the investigation window. A GDPR Data Processing Agreement is available on the Enterprise plan.
Do I need to configure my CloudWatch alarms differently? No. ConvOps connects to your existing CloudWatch alarms via SNS. No changes to alarm thresholds, namespaces, or configurations required.
Can I disconnect ConvOps at any time? Yes. Delete the IAM role from your AWS console and ConvOps immediately loses all access. No support ticket, no waiting period.

Ready to stop debugging alerts manually?

Connect your AWS account in 2 minutes. No credit card.

Get Started Free