Security 101

What Is MDR With Custom Detection Rules?

Custom detection rules are written for your environment and kept in your own Sentinel workspace, where your team can read and change them.

MDR with custom detection rules means the provider writes and tunes detections for your environment, and the rules live where your team can read and change them. In a Microsoft environment, that usually means analytics rules in your own Sentinel workspace.

Accelerynt MDR works this way. Our team builds detection rules directly in your Sentinel workspace, and your team can read and change every one of them.

Many managed services run one detection set across every customer. That keeps the service consistent, and it can leave your team unable to adjust a rule that does not fit how your environment works.

Custom rules, defined

  • Written for your environment. Rules reflect what normal activity looks like for your organization.
  • Stored in your workspace. The rule logic sits in your Sentinel, where your team can read it.
  • Open to your team. Your analysts can change a rule or write their own alongside the provider’s.
  • Tuned with your input. Your team’s knowledge of normal activity shapes what each rule flags.

How do detection rules work in Microsoft Sentinel?

Most Sentinel detections are scheduled analytics rules. Each one runs a Kusto Query Language (KQL) query at a set interval and looks back over a defined window of data.

5 min to 14 days

The allowed range for a scheduled rule’s query interval and its lookback period in Microsoft Sentinel. Source: Microsoft Learn, scheduled analytics rules.

Rules start in one of two places. Templates are written by Microsoft or by the vendor of a connected solution, and you can use them as written or customize them. Custom rules are built from scratch with your own query.

Source: Microsoft Learn, create a scheduled analytics rule.

Global detections or rules in your workspace?

The difference shows up the first time a rule fires on something your team knows is normal.

QuestionOne detection set for every customerRules in your Sentinel workspace
Where does the rule logic live?In the provider’s platformIn your own Sentinel workspace
Can your team read the query?It depends on the providerYes, every rule is visible in your tenant
Can your team change a rule?Often only through the providerYes, and your team can write its own
How is it tuned?Across the provider’s whole customer baseWith your team’s knowledge of normal activity
Tuning approach from the Accelerynt MDR FAQs.

How does a custom detection rule get built?

  1. Name the behavior. Start from an activity worth flagging in your environment, such as an admin account signing in from a location it has never used.
  2. Write the query. Express that behavior as a KQL query against the tables your data connectors fill.
  3. Set the schedule and lookback. Choose how often the rule runs and how far back it reads, within Sentinel’s limits.
  4. Test against your own data. Run the query against recent logs to see what it would have flagged and how often.
  5. Tune with your team. Adjust thresholds and exclusions using what your team knows about normal activity, and record each change.

For how two teams share one workspace day to day, read what co-managed Microsoft Sentinel means.

What should you ask an MDR provider about custom rules?

  • Ask where the rules live. Confirm they sit in your own workspace and stay there.
  • Ask to read one. A provider confident in its rules can show you the query behind an alert.
  • Ask who can change them. Confirm your team can edit a rule or add its own without waiting on a ticket.
  • Ask how tuning is recorded. Each change should be visible in your tenant so your team can review it.

Accelerynt MDR builds every rule directly in your Sentinel workspace, and your team has full access to each one and can keep writing its own.

For how noisy rules get fixed at the source, read how managed security providers reduce alert fatigue. For what else you should be able to see, read which MDR providers let you see what their SOC is doing.

How Accelerynt handles detection rules

Accelerynt builds detection rules directly in your Sentinel workspace, connected through Azure Lighthouse. Your team has full access to every rule we build and can keep writing its own.

We adjust detection logic based on your team’s knowledge of normal activity, and our playbooks are public on GitHub. We guarantee that under 1% of what we escalate to your team is noise or a false positive.

See Managed Detection and Response and how we work together inside your tenant. For why more alerts rarely means more security, read our post Seeing Everything. Acting on Nothing.

Custom rules matter only once you have decided to bring in an MDR. If that decision is still open, read whether Microsoft Defender is enough or you need a separate MDR.

Custom rules still need tuning once they meet real activity in your tenant. Read how to tune Microsoft Sentinel detection rules for the steps.

Frequently asked questions

Is there an MDR that lets us write our own detection rules?

Yes. Accelerynt builds detection rules directly in your Sentinel, and your team keeps full access to every rule and can write its own. In general, look for a provider whose rules live in your own Sentinel workspace.

What is a custom detection rule in Microsoft Sentinel?

It is a scheduled analytics rule built from scratch with your own KQL query, instead of starting from a Microsoft or vendor template. Sentinel runs it on an interval you set and reads back over a lookback window, each between 5 minutes and 14 days.

Can an MDR provider tune detections for our environment?

Yes, when the rules run in your workspace. Accelerynt tunes your Sentinel analytics rules with your team’s input on normal activity and builds suppression for false positives at the source.

Who owns detection rules an MDR provider builds?

When the rules are built in your own Sentinel workspace, they stay in your tenant. Accelerynt builds everything in your workspace, so every rule and playbook stays yours.

Talk to a Microsoft security engineer

We work inside your Microsoft environment, with your team, and show you where to focus first.