Under attack?
Software engineering

Extension development
for your SIEM

Your vendor’s catalogue covers the most common sources. For everything else — specialised devices, line-of-business applications, local sources — it has to be built. That is what we do, across every platform in our scope.

  • Collection connectors
  • Parsers & normalisation
  • Dashboards
  • Detection use cases
  • Packaging & publication
The need

An unnormalised source is a half-useless source

Ingesting raw logs into a SIEM is not enough. Without field extraction, without mapping to a common data model, a source can feed neither the correlation rules, nor the dashboards, nor the compliance reports. It consumes licence and produces no detection.

A well-built extension solves that once and for all: it turns a proprietary feed into data the whole platform can use — rules, dashboards and reports included. Each vendor has its own mechanics; the principle stays the same.

Public reference · Splunkbase

Microsoft Windows Firewall Observability

A Splunk application built by our teams and published on Splunkbase, filling a gap in the catalogue: giving real operational visibility over Windows Firewall logs, often collected but rarely used.

Event extraction and normalisation, dashboards tracking rules and blocks, detection of configuration changes.

Our deliverables

What an Avangard extension contains

Collection and parsing

Sourcetype definitions, field extraction, handling of multi-line formats and timestamps, event breaking.

Normalisation

Mapping to the platform’s data model — CIM on Splunk, ECS on Elastic, ASIM on Sentinel — the precondition for feeding the standard rules.

Dashboards

Ready-to-use operational views, designed to be readable by an analyst who does not know the source.

Detection use cases

Correlation searches and alerts shipped with the application, documented and mapped to MITRE ATT&CK techniques.

Documentation

Installation and configuration guide, description of extracted fields, prerequisites, version compatibility matrix.

Packaging and publication

Validation through the vendor’s inspection tooling, alignment with its requirements, publication on the official catalogue where it makes sense, and version tracking.

By platform

Every vendor has its own vocabulary

The principle is identical — making a source usable — but the mechanisms and formats differ. We work across every platform in our scope.

PlatformWhat we buildData model
SplunkApplications and technology add-ons, publishable on SplunkbaseCIM
Microsoft SentinelData connectors, parsing functions, analytics rules and workbooksASIM
Elastic SecurityIntegrations, ingest pipelines, processors and dashboardsECS
OpenText ArcSightConnectors, notably FlexConnectors for proprietary formatsCEF
WazuhDecoders, detection rules and integration modulesWazuh schema
The case of local sources

Line-of-business applications built in house, devices from regional vendors, legacy systems whose format is documented nowhere: these are precisely the sources no vendor catalogue will ever cover, and often the richest in signal for your context.

Two modes

Private or published extension

Private extension

Built for your use alone and deployed on your platform. You own the code and are free to develop it further.

  • Internal sources and proprietary business applications
  • Specific detection logic you would rather not expose
  • Confidentiality constraints on data schemas

Extension published to the catalogue

Made available to the community under our name or yours, with maintenance and compatibility assured over time.

  • Visibility across the global Splunk ecosystem
  • Vendor positioning for your organisation
  • Sources of general interest, not specific to one client
The sequence

From log sample to delivered extension

1

Analysis

Study of real samples and of the source documentation.

2

Design

Target data model, fields to extract, mapping to the platform schema.

3

Development

Configuration, extractions, views and saved searches.

4

Validation

Testing on real data, vendor checks, performance review.

5

Delivery

Deployment, documentation, publication and version tracking.

A source your SIEM ignores?

Send us a log sample and the source documentation. We will tell you quickly what can be drawn from it and at what effort.