# Monitoring

URL: https://donedeploy.com/devops-services/monitoring/

observability · prod

Requests/s**2,140**

Errors**0.4%**

p95**412 ms**

CPU**61%**

checkout p95 latency**412 ms**

anomaly

✓

**Likely cause**slow query after deploy v2.14

found

**Grouping alerts**14 alerts into 1 incident

running

*AI*

**Root cause hint**in 40 seconds

*✓*

**1 incident**not 14 alerts

How we work

## From design to implementation, and everything after

**Design**the service, its architecture and costs

**Build**infrastructure and code, reviewed

**Run**deployed, secured and watched

**Automate**pipelines, runbooks and AI

**Improve**cost, speed and reliability, every month

The problem

## What goes wrong without it

1

### Alert fatigue

So many alerts fire that the important ones get ignored.

2

### Blind spots

Some services have no metrics or traces at all, so failures there are invisible until users complain.

3

### Slow investigations

Finding the cause of an incident means jumping between tools and reading raw logs by hand.

### What we do

- Instrument services with OpenTelemetry for metrics, logs and traces
- Set up Prometheus and Grafana, or work with Datadog and Sentry if you already use them
- Define alerts around user-facing symptoms rather than every internal number
- Use AI to group related alerts and write a plain summary of what is happening
- Use AI to draft incident timelines and post-incident reports for engineers to check
- Monitor AI and LLM services too: latency, errors, token use and cost

### What you get

- Dashboards for each service and for the business-critical paths
- Alert rules and routing in code
- On-call and incident runbooks
- A post-incident report template with AI-drafted summaries

How it works

## From first call to continuous improvement

### Free first consultation

We learn what you monitor today and which failures hurt most.

### Review

We review existing dashboards, alerts and gaps, with AI helping sift through alert history.

### Fixed-price proposal

We agree the services to cover, the tools and the price.

### Build

Engineers instrument services, build dashboards and alerts, and add AI summaries.

### Improve after launch

We review alert quality and incidents regularly and tune what is noisy or missing.

Technologies

## What we use

[!OpenTelemetry](https://donedeploy.com/technologies/opentelemetry/)[!Prometheus](https://donedeploy.com/technologies/prometheus/)[!Grafana](https://donedeploy.com/technologies/grafana/)[!Datadog](https://donedeploy.com/technologies/datadog/)[!Sentry](https://donedeploy.com/technologies/sentry/)[!Elasticsearch](https://donedeploy.com/technologies/elasticsearch/)[!Kubernetes](https://donedeploy.com/technologies/kubernetes/)

From the blog

## Related articles

[!AI · October 6, 2026Observability for LLM ApplicationsWhat to measure in an LLM application: tokens, latency, cost per request, OpenTelemetry traces and the quality signals that show when answers get worse.](https://donedeploy.com/llm-observability/)[!Devops · October 1, 2024Serverless MonitoringLearn the key metrics and tools for serverless monitoring and observability, including AWS CloudWatch and Datadog, to ensure optimal performance](https://donedeploy.com/serverless-monitoring/)

Related services

## Often done together

[**After launch**Infrastructure that keeps getting cheaper, faster and safer after launch*Learn more »*](https://donedeploy.com/devops-services/continuous-improvement/)[**AI service design**Put AI features into production with predictable cost and control*Learn more »*](https://donedeploy.com/devops-services/ai-service-design/)[**Containers and orchestration**Containers and Kubernetes sized for your team, not for a giant*Learn more »*](https://donedeploy.com/devops-services/containers-and-orchestration/)

Questions

## Frequently asked

Does AI decide what to do during an incident?

No. AI summarises and suggests likely causes; engineers decide and act.

Do we have to replace our current tools?

No. We can work with Datadog, Sentry or an existing Grafana setup, and use open-source tools where they fit better.

Can you monitor our LLM features?

Yes. We track latency, errors, token use and cost per feature so you can see how AI services behave in production.

Tell us what you run and what you want to change. We reply personally, and the first consultation is free.

[Let's talk »](https://donedeploy.com/lets-talk/)