CrowdStrike Next-Gen SIEM
Overview
CrowdStrike Falcon Next-Gen SIEM — built on the former Falcon LogScale (Humio) engine — ingests third-party data over a plain HTTPS endpoint called the HEC (HTTP Event Collector), modeled on Splunk’s HEC convention. Britive can push its audit log to any HTTPS URL through a webhook notification medium. Pointing one at the other gives you a live feed of Britive audit activity in Next-Gen SIEM with nothing to install.
What you’ll accomplish:
- Create an HEC data connector in Falcon Next-Gen SIEM and get its API key and ingest URL
- Write a custom parser so Britive’s audit JSON lands as structured, searchable fields instead of a raw blob
- Confirm the connector accepts a test event
- Configure a webhook notification medium in Britive pointed at it
- Attach the notification medium to Britive’s audit log so matching events are forwarded automatically
Prerequisites
Before you begin, make sure you have:
- A Britive tenant with System admin access (to configure notification mediums and audit log webhooks)
- A Falcon Next-Gen SIEM or Next-Gen SIEM 10GB subscription
- Administrator or Connector Manager access to the Falcon console
- Your CrowdStrike cloud, since Next-Gen SIEM is only available on some of them: US-1, US-2, EU-1, or US-GOV-1
Never hardcode the HEC API key in code or commit it to a repository. It’s shown only once, at connector creation — store it in a secrets manager immediately.
How It Works
graph LR
A["Britive Audit Log"] -->|"event matches<br/>filter expression"| B["Webhook<br/>Notification Medium"]
B -->|"HTTPS POST<br/>Authorization: Bearer api_key"| C["HEC Data Connector<br/>/services/collector/raw"]
C --> D["Custom CQL Parser"]
D --> E["Structured event in<br/>Advanced Event Search"]
| Stage | What happens |
|---|---|
| Filter | Britive evaluates every audit event against the filter expression on the audit log webhook |
| Forward | A matching event is POSTed as raw JSON — Britive’s own audit record, with no envelope wrapper — to the URL configured on the webhook notification medium |
| Authenticate | The HEC connector validates the Authorization header on every request |
| Parse | The parser assigned to the connector runs on each event, extracting fields and a timestamp into Next-Gen SIEM’s schema |
| Index | The parsed event becomes searchable in Advanced Event Search, taggable by the connector’s #connector tag |
Creating the connector is a one-time setup step, done before any events can flow — the connector is also where the parser assignment lives, which determines whether incoming events arrive as structured, searchable fields or get left as a raw blob.
Create the HEC Data Connector
Open Data Connectors
In the Falcon console, go to Next-Gen SIEM → Data ingestion → Data connectors. Click Add data connector.
Select the HEC/HTTP Event Data Connector
Find and select HEC/HTTP Event Data Connector. In the New connection dialog, review the connector metadata and click Configure.
Configure the connector
Enter a Connector name and optional Description. For Parsers, select an existing parser or your own custom one — see Write a Custom Parser below, since CrowdStrike has no built-in Britive parser. Accept the Terms and Conditions and click Save.
Copy the API key and URL
Open the connector’s detail page. Copy the API key and URL immediately — the API key is shown only once.
Build the Ingestion Endpoint and Headers
The endpoint is built from your connector’s URL key and your CrowdStrike cloud’s ingest host, and must end in /raw:
https://<connector-url-key>.<ingest-host>/services/collector/rawUse /services/collector/raw, not /services/collector. The structured endpoint (no /raw) expects a Splunk-HEC-style envelope — {"event": {...}, "fields": {...}} — around every event. Britive’s webhook has no way to wrap its payload like that; it POSTs its own audit JSON directly as the request body. Pointing the webhook at the non-raw endpoint is the single most common cause of this integration silently not working — events either don’t show up at all, or arrive as garbled, incomplete @rawstring content that the parser can’t make sense of.
| CrowdStrike cloud | Ingest host |
|---|---|
| US-1 | ingest.us-1.crowdstrike.com |
| US-2 | ingest.us-2.crowdstrike.com |
| EU-1 | ingest.eu-1.crowdstrike.com |
| US-GOV-1 | ingest.laggar.gcw.crowdstrike.com |
The connector’s URL key is an opaque identifier (not something you choose) shown on the connector’s detail page — for example 4ab0a7ab8351c9.ingest.us-1.crowdstrike.com.
Every request needs these headers:
Authorization: Bearer <api_key>
Content-Type: application/jsonWrite a Custom Parser
CrowdStrike parsers are written in CQL (CrowdStrike Query Language), following the CrowdStrike Parsing Standard (CPS), which builds on the Elastic Common Schema.
CrowdStrike’s parser-authoring guidelines assume a parser receives one event at a time, with no extra transport wrapper around the vendor’s native JSON. Britive’s webhook delivers exactly that — one audit event, its own native JSON, per HTTP request — which is exactly what the /raw endpoint above is for.
What Britive actually sends
This is a real event captured from a Britive audit webhook, with placeholder values standing in for real tenant data — the field names and nesting are exactly as sent:
{
"timestamp": "2026-10-05T19:45:28.651+0000",
"id": "",
"actor": {
"userId": "cjbddqc73pbdm8lvkp7b",
"displayName": "Jane Doe",
"type": "User",
"authenticationType": "Bearer",
"role": null,
"username": "jane.doe@example.com",
"impersonatedBy": null,
"credential": {
"id": null,
"name": "jane.doe@example.com",
"type": "Bearer",
"idpType": null,
"idpName": null
}
},
"client": {
"device": "Desktop",
"ipAddress": "203.0.113.10",
"userAgent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/154.0.0.0 Safari/537.36",
"displayName": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/154.0.0.0 Safari/537.36 - 203.0.113.10",
"browser": "Edge",
"platform": "Win10",
"asn": "64512",
"country": "US",
"countryRegion": "CA",
"latitude": "37.73080",
"longitude": "-122.38380",
"metroCode": "807",
"additionalInfo": {},
"XForwardedFor": null
},
"event": {
"eventType": "access.checkout.request",
"displayName": "Profile Check Out Request",
"additionalInfo": {}
},
"target": {
"targetId": "brn5uxd67asoml99oc4s",
"applicationName": "AWS Production",
"environmentName": "111111111111 (Acme Corp)",
"environmentGroupName": null,
"parentEnvironmentGroupName": null,
"displayName": "AWS Production | 111111111111 (Acme Corp) | Developer",
"applicationSessionId": "janedoe-q25skcy7p4gapv",
"email": null,
"additionalInfo": { "accessType": "CONSOLE", "justification": null }
},
"result": { "message": null, "success": true },
"tenantId": "demo"
}A few things confirmed from this shape:
idis always an empty string. Britive doesn’t issue a per-event identifier.timestampcarries an explicit+0000(UTC) offset, but it’s an ISO 8601 string —@timestampinternally needs epoch milliseconds, so it can’t be assigned directly (more on this below).
Handling events with no client context
Some Britive audit events are generated by Britive’s own automation rather than an interacting human — actor.type shows "System User" on these. Because there’s no browser session behind them, every client.* field (IP address, user agent, geo, and so on) is JSON null.
parseJson() keeps JSON nulls as the literal string "null" by default. If a parser maps client.ipAddress straight into ECS’s source.ip field, Next-Gen SIEM rejects the whole event with a schema violation (source.ip must be a valid IPv4/IPv6 address, and the string "null" isn’t one). The same applies to any other type-checked ECS field you map from a client.* value — geo coordinates and ASN included, if you add them later.
Two changes fix it:
- Discard nulls and empty values at parse time:
parseJson(prefix="Vendor.", excludeEmpty=true, handleNull="discard"). This also cleans up the emptyidfield on every event. - Only map
source.ipanduser_agent.originalwhen Britive actually supplied a value — guard each with acase{}block rather than mapping unconditionally.
The parser
parseJson(prefix="Vendor.", excludeEmpty=true, handleNull="discard")
| parseTimestamp(format="yyyy-MM-dd'T'HH:mm:ss.SSSZ", field=Vendor.timestamp)
| Cps.version := "1.2.0"
| ecs.version := "8.11.0"
| Parser.version := "1.2.0"
| observer.vendor := "Britive"
| observer.product := "Britive"
| event.kind := "event"
| event.module := "britive"
| event.category[0] := "iam"
| event.action := Vendor.event.eventType
| case {
Vendor.event.eventType = /checkin/i
| event.type[0] := "end";
Vendor.event.eventType = /checkout/i
| event.type[0] := "start";
*
| event.type[0] := "info";
}
| case {
Vendor.result.success = "true"
| event.outcome := "success";
Vendor.result.success = "false"
| event.outcome := "failure";
*
| event.outcome := "unknown";
}
| user.id := Vendor.actor.userId
| user.name := Vendor.actor.username
| user.full_name := Vendor.actor.displayName
| case {
Vendor.client.ipAddress = * | Vendor.client.ipAddress != "null"
| source.ip := Vendor.client.ipAddress;
*
;
}
| case {
Vendor.client.userAgent = * | Vendor.client.userAgent != "null"
| user_agent.original := Vendor.client.userAgent;
*
;
}Configure the Webhook Notification Medium in Britive
Open Notification Mediums
Log in to Britive with administrator privileges and go to System admin → Global Settings → Notification Mediums.
Add a medium
Click Add Medium. Enter a Name (for example CrowdStrike Next-Gen SIEM) and an optional Description.
Select Webhook
From the application drop-down, select Webhook. Enter the HEC endpoint URL you built above — ending in /raw — as the Webhook URL.
Add the headers
Under request headers, add both:
| Header | Value |
|---|---|
Authorization | Bearer <api_key> |
Content-Type | application/json |
Save
Click Save. This medium is now available to attach to the audit log.
If you edit this medium’s URL or headers later, delete and re-add the corresponding entry under Manage Webhooks — updating the notification medium alone does not retroactively apply to a webhook entry that was already attached to the audit log. This is the single most common reason a fix “doesn’t seem to take.”
Forward Audit Logs to the Webhook
Open Manage Webhooks
Go to System admin → Audit Log → Manage Webhooks.
Add the webhook
Click Add Webhook. Select the CrowdStrike Next-Gen SIEM notification medium you just created.
Set the filter expression
Leave the filter blank to receive all audit logs, or add a filter to receive a subset.
The filter expression field uses a function-call syntax — operator(field, 'value') — matching what the Manage Webhooks screen itself shows as an example:
| Filter | Keeps |
|---|---|
contains(event.eventType, 'access.checkout.request') | Checkout request events — the example shown directly on the Manage Webhooks screen |
Only the contains(...) form above is confirmed against a live tenant. The Audit Log’s Advanced Search also documents STARTS WITH, NOT EQUALS, and EQUALS operators, which likely correspond to similarly-named functions here — and multiple conditions likely combine with and, based on the same page. Confirm the exact function names and how to combine conditions directly on your tenant’s Manage Webhooks screen before building anything beyond a single contains(...) filter.
Save
Click Save.
Verify
Confirm the connector is active
Go to Next-Gen SIEM → Data ingestion → Data connectors and confirm the connector’s Status shows Active.
Generate a real Britive event
Perform an action in Britive that matches your filter.
Confirm the event parses cleanly
In the connector’s parser editor (or Advanced Event Search), confirm event.action, user.name, user.full_name, and event.outcome are populated, @timestamp matches when the action actually happened in Britive (not ingestion time), and there are zero schema violations.
Confirm a System User event parses cleanly too
Trigger or find an event where actor.type is "System User" and confirm it also shows zero schema violations, with source.ip absent (not present-and-null) on that event — this is what confirms the null-handling guard in the parser is working.
Troubleshoot
| Symptom | Likely Cause | Fix |
|---|---|---|
Connector shows idle, or events arrive as garbled/incomplete @rawstring content | Webhook URL points at /services/collector instead of /services/collector/raw | Britive posts raw audit JSON with no envelope — the URL must end in /raw. See Build the Ingestion Endpoint and Headers |
| Events still don’t improve after fixing the medium’s URL | The existing entry under Manage Webhooks doesn’t pick up a medium URL change on its own | Delete the webhook entry under Manage Webhooks and add it again |
Schema violation: “The given IP address is not a valid IPv4 or IPv6 address” on source.ip | A System User event has client.ipAddress: null, and parseJson() turns JSON null into the literal string "null" by default | Use excludeEmpty=true, handleNull="discard" on parseJson(), and only map source.ip/user_agent.original inside a case{} guard that excludes "null" — both are in the parser above |
@id/id or @timestamp/timestamp “near duplicate field” schema violations | parseJson() called without a prefix, so raw fields collide with reserved Next-Gen SIEM fields | Use parseJson(prefix="Vendor.", ...) |
| Timestamps show ingest time instead of event time | @timestamp wasn’t set via parseTimestamp(), or it was assigned the raw ISO string directly | @timestamp expects epoch milliseconds, not a string — use parseTimestamp(format="yyyy-MM-dd'T'HH:mm:ss.SSSZ", field=Vendor.timestamp) |
event.outcome always shows "unknown" | Vendor.result.success compared as an unquoted boolean instead of the string "true"/"false" | parseJson() flattens booleans to text — compare against the quoted strings |
user.name shows systemuser on some events | Expected — some audit events are generated by Britive’s own automation and logged under actor.type: "System User" rather than an interacting human | None needed — this is normal for automated/backend actions |
Test request returns 401/403 | Missing or incorrect Authorization header | Re-check the API key copied from the connector’s detail page |
| Notification medium can’t be selected under Manage Webhooks | Medium wasn’t saved as type Webhook, or wasn’t saved at all | Re-check the medium under Global Settings → Notification Mediums |