Open source cloud-native security lake platform (SIEM alternative) for threat hunting, detection & response, and cybersecurity analytics at petabyte scale on AWS

Open issues
Closed issues
Last commit
6 months ago
Total releases
Total commits
Open PRs
Closed PRs
Repo URL
Offers premium version?

Matano is an open source security lake platform for AWS

πŸ”” Read our announcement blog post πŸ””

Matano is an open source cloud-native alternative to SIEM, built for security teams on AWS.

We are on a mission to build the first open platform for threat hunting, detection & response, and cybersecurity analytics at petabyte scale.

    <a href="" rel="nofollow">Docs</a>
    <a href="" rel="nofollow">Website</a>
    <a href="" rel="nofollow">Community</a>


  • Security Data Lake: Matano normalizes unstructured security logs into a structured realtime data lake in your AWS account.
  • Collect All Your Logs: Matano integrates out of the box with 50+ sources for security logs and can easily be extended with custom sources.
  • Detection-as-Code: Use Python to build realtime detections as code. Support for automatic import of Sigma detections to Matano.
  • Log Transformation Pipeline: Matano supports custom VRL (Vector Remap Language) scripting to parse, enrich, normalize and transform your logs as they are ingested without managing any servers.
  • No Vendor Lock-In: Matano uses an open table format (Apache Iceberg) and open schema standards (ECS), to give you full ownership of your security data in a vendor-neutral format.
  • Bring Your Own Analytics: Query your security lake directly from any Iceberg-compatible engine (AWS Athena, Snowflake, Spark, Trino etc.) without having to copy data around.
  • Serverless: Matano is fully serverless and designed specifically for AWS and focuses on enabling high scale, low cost, and zero-ops.


πŸ‘€ Use cases

  • Reduce SIEM costs drastically (1/10th the cost).
  • Augment your SIEM with a security data lake for additional context during investigations.
  • Instantly search for matches to IOCs across your data lake using standardized fields.
  • Write detections-as-code using Python to detect suspicious behavior & create contextualized alerts.
  • Easier to use cloud-native open source SIEM alternative for detection & response.
  • ECS-compatible serverless alternative to ELK / Elastic Security stack.

✨ Integrations Managed log sources

Alert destinations

Enrichment / Threat Intelligence

Query engines

Quick start

View the complete installation instructions


Install the matano CLI to deploy Matano into your AWS account, and manage your Matano deployment.


curl -OL
chmod +x
sudo ./


curl -OL
chmod +x
sudo ./

Read the complete docs on getting started

To get started with Matano, run the matano init command.

  • Make sure you have AWS credentials in your environment (or in an AWS CLI profile).
  • The interactive CLI wizard will walk you through getting started by generating an initial Matano directory for you, initializing your AWS account, and deploying Matano into your AWS account.
  • Initial deployment takes a few minutes.

Directory structure

Once initialized, your Matano directory is used to control & manage all resources in your project e.g. log sources, detections, and other configuration. It is structured as follows:

➜  example-matano-dir git:(main) tree
β”œβ”€β”€ detections
β”‚   └── aws_root_credentials
β”‚       β”œβ”€β”€
β”‚       └── detection.yml
β”œβ”€β”€ log_sources
β”‚   β”œβ”€β”€ cloudtrail
β”‚   β”‚   β”œβ”€β”€ log_source.yml
β”‚   β”‚   └── tables
β”‚   β”‚       └── default.yml
β”‚   └── zeek
β”‚       β”œβ”€β”€ log_source.yml
β”‚       └── tables
β”‚           └── dns.yml
β”œβ”€β”€ matano.config.yml
└── matano.context.json

When onboarding a new log source or authoring a detection, run matano deploy from anywhere in your project to deploy the changes to your account.

πŸ”§ Log Transformation & Data Normalization

Read the complete docs on configuring custom log sources

Matano uses Vector Remap Language (VRL), to allow users to easily onboard custom log sources and encourages you to normalize fields according to the Elastic Common Schema (ECS) to enable enhanced pivoting and bulk search for IOCs across your security data lake.

Users can define custom VRL programs to parse and transform unstructured logs as they are being ingested through one of the supported mechanisms for a log source (e.g. S3, SQS).

VRL is an expression-oriented language designed for transforming observability data (e.g. logs) in a safe and performant manner. It features a simple syntax and a rich set of built-in functions tailored specifically to observability use cases.

Example: parsing JSON

Let's have a look at a simple example. Imagine that you're working with HTTP log events that look like this:

  "line": "{\"status\":200,\"srcIpAddress\":\"\",\"message\":\"SUCCESS\",\"username\":\"ub40fan4life\"}"

You want to apply these changes to each event:

  • Parse the raw line string into JSON, and explode the fields to the top level
  • Rename srcIpAddress to the source.ip ECS field
  • Remove the username field
  • Convert the message to lowercase

Adding this VRL program to your log source as a transform step would accomplish all of that:


transform: |
  . = object!(parse_json!(string!(.json.line)))
  .source.ip = del(.srcIpAddress)
  .message = downcase(string!(.message))

schema: ecs_field_names: - source.ip - http.status

The resulting event πŸŽ‰:

  "message": "success",
  "status": 200,
  "source": {
    "ip": ""
πŸ“ Writing Detections

Read the complete docs on detections

Use Matano detections to define rules that can alert on threats in your security logs. Matano users define detections as code (DaC). A detection is a Python program that is invoked with data from a log source in realtime and can create an alert.

Examples Detect failed attempts to export AWS EC2 instance in AWS CloudTrail logs.

def detect(record):
  return (
    record.deepget("event.action") == "CreateInstanceExportTask"
    and record.deepget("event.provider") == ""
    and record.deepget("event.outcome") == "failure"
Detect Brute Force Logins by IP across all configured log sources (e.g. Okta, AWS, GWorkspace)
def detect(r):
    return (
        "authentication" in r.deepget("event.category", [])
        and r.deepget("event.outcome") == "failure"

def title(r): return f"Multiple failed logins from {r.deepget('user.full_name')} - {r.deepget('source.ip')}"

def dedupe(r): return r.deepget("source.ip")


  • aws_cloudtrail
  • okta_system
  • o365_audit alert: severity: medium threshold: 5 deduplication_window_minutes: 15 destinations:
    • slack_my_team
Detect Successful Login from never before seen IP for User
from detection import remotecache

a cache of user -> ip[]

user_to_ips = remotecache("user_ip")

def detect(record): if ( record.deepget("event.action") == "ConsoleLogin" and record.deepget("event.outcome") == "success" ): # A unique key on the user name user = record.deepget("")

    existing_ips = user_to_ips[user] or []
    updated_ips = user_to_ips.add_to_string_set(

    # Alert on new IPs
    new_ips = set(updated_ips) - set(existing_ips)
    if existing_ips and new_ips:
        return True</pre> 

🚨 Alerting

Read the complete docs on alerting

Alerts Matano table

All alerts are automatically stored in a Matano table named matano_alerts. The alerts and rule matches are normalized to ECS and contain context about the original event that triggered the rule match, along with the alert and rule data.

Example Queries

Summarize alerts in the last week that are activated (exceeded the threshold)

select as alert_id, as rule_name,
  max(matano.alert.title) as title,
  count(*) as match_count,
  min(matano.alert.first_matched_at) as first_matched_at,
  max(ts) as last_matched_at,
  array_distinct(flatten(array_agg(related.ip))) as related_ip,
  array_distinct(flatten(array_agg(related.user))) as related_user,
  array_distinct(flatten(array_agg(related.hosts))) as related_hosts,
  array_distinct(flatten(array_agg(related.hash))) as related_hash
  matano.alert.first_matched_at > (current_timestamp - interval '7' day)
  and matano.alert.activated = true
group by,
order by
  last_matched_at desc
Delivering alerts

Matano allows you to deliver alerts to external systems. You can use the Matano alerting SNS topic to deliver alerts to Email, Slack, and other services.

A medium severity alert delivered to Slack

❔ Why Matano?

  • Traditional tools used to analyze security data (SIEMs) don’t scale, and are too expensive and difficult to manage for cloud-based security teams.
  • Cybersecurity vendors lock your data in proprietary formats which make it difficult to use outside of their product. With Matano, all your data is in open Apache Iceberg tables that can can be directly queried from different tools (AWS Athena, Snowflake, etc.) without having to copy any data.
  • Security is a Big Data problem: collecting data from your network, SaaS, and cloud environments can exceed 100TBs of data. Security teams are forced to either not collect some data, leave data unprocessed, or build an in-house data lake to cost-effectively analyze large datasets. Matano helps you easily build a security data lake with all features needed for detection and response.
  • At scale, without a strategy to normalize data into a structured format, it is difficult to correlate across data sources & build effective alerts that don’t create many false positives. Traditional SIEM query-based rules fail to accurately identify threats. Matano's detection-as-code approach offers greater flexibility and help's you harden your detections over time.

❀️ Community support

For general help on using Matano, please refer to the official Matano documentation. For additional help, feel free to use one of these channels to ask a question:

  • Discord (Come join the Matano family, and hang out with the team and community)
  • Forum (For deeper conversations about features, the project, or problems)
  • GitHub (Bug reports, Contributions)
  • Twitter (Get news hot off the press)

πŸ‘· Contributors

Thanks go to these wonderful people (emoji key):

Shaeq Ahmed




Kai Herrera

πŸ’» πŸ€” πŸš‡


πŸ› πŸ€” πŸ““

Zach Mowrey

πŸ€” πŸ› πŸ““


πŸ““ πŸ› πŸ€”

Greg Rapp

πŸ› πŸ€”

Matthew X. Economou


Jarret Raim


Matt Franz


Francesco Faenzi


Nishant Das Patnaik


Tim O'Guin

πŸ€” πŸ› πŸ’»

Francesco R.


Joshua Sorenson

πŸ’» πŸ“–

This project follows the all-contributors specification. Contributions of any kind are welcome!


Alternative Projects

Subscribe to Open Source Businees Newsletter

Twice a month we will interview people behind open source businesses. We will talk about how they are building a business on top of open source projects.

We'll never share your email with anyone else.