Capabilities

Everything you need to collect, shape, route, and govern telemetry and event data at the first mile

LyftData exposes seven core capabilities — Collect, Transform, Route, Compose, Scale, Observe, Visual Editor — all running on the deterministic Server → Workers → Jobs architecture. Each capability builds toward control, scale, and governance.

Works with logs, metrics, traces, security events, SaaS exports, files, and API records.

See Solutions →Explore the Docs →

Here’s what you can do with LyftData today.

Three themes

To keep complexity understandable, every primitive belongs to one of these themes:

CONTROL & COST

  • •Shape and mask data before it hits metered tools
  • •Route curated subsets into expensive platforms
  • •Archive full fidelity cheaply

Every byte is intentional, not accidental.

SCALE & RELIABILITY

  • •Server → Workers → Jobs keeps behavior deterministic
  • •Workers scale horizontally without rewrites
  • •Run & Trace shows exactly what will happen before deployment

Pipelines stay predictable even as volume grows.

GOVERNANCE & AUDITABILITY

  • •Jobs are signed, version-controlled definitions
  • •Run & Trace produces auditable evidence
  • •Server stores lineage and change history

You can prove how data moved — every time.

Collect

Workers read directly from your existing sources with a single Input definition per Job—for telemetry, events, and records.

  • Files & file shares

  • S3 / GCS / Azure Blob

  • Windows Events

  • HTTP / APIs

  • Security sources like CrowdStrike

What it solves

  • Tool-specific agents and scripts

  • Duplicated ingestion logic

  • Missed sources due to inconsistent configs

A consistent, governed ingestion layer.

Inputs & Sources →

Transform (Actions)

Chain declarative Actions to filter, parse, enrich, mask, script, and normalize events and records.

  • Filter — remove noise

  • Parse — extract fields

  • Enrich — add context

  • Redact / Mask — govern sensitive data

  • Script — run custom logic

  • Normalize — produce consistent structures

What it solves

  • Brittle per-tool transforms

  • Untraceable scripts

  • Inconsistent masking and enrichment

Cleaner, reviewable data upstream.

Actions Reference →

Route

Each Job defines the primary Output so you can intentionally deliver to SIEM, observability, analytics, or archives—without vendor lock-in.

  • SIEM

  • Observability tools

  • Data platforms

  • Object storage

  • Archival buckets

What it solves

  • Over-ingesting into expensive tools

  • Blind forwarding

  • Duplicated pipelines per destination

Every destination receives only the data it needs.

Outputs →

Compose (Channels)

Channels clone and fan out governed outputs so one Job can feed many downstream systems—from telemetry tools to data platforms.

  • Curated logs → SIEM

  • Structured events → observability

  • Full-fidelity logs → archive

What it solves

  • Copy-paste pipelines per tool

  • Inconsistent policies across branches

  • Costly re-ingestion when adding destinations

One pipeline, many intentional outputs.

Channels tutorial →

Scale

Workers are stateless executors you can add anywhere without rewriting Jobs.

What it solves

  • Per-cluster drift

  • Scaling tied to agents

  • Risky deployments under load

Predictable performance and safer parallelism.

Operate & Scale →

Observe

LyftData surfaces logs, metrics, and traces from Workers back to Server.

What it solves

  • Blind pipelines

  • Slow failure detection

  • Unverifiable transformations

Observability by default.

Operate Overview →

Visual Editor

Author Jobs visually, preview transformations with Run & Trace, then export or review in YAML.

What it solves

  • Hidden UI-only logic

  • Slow pipeline authoring

  • Handwritten configs without guardrails

Speed during creation, precision during review.

Visual Editor →

Workflows

Workflows & deployments

Versioned releases, safe rollouts, and clear rollbacks—across one worker or a fleet.

Build

  • Draft workflows quickly and iterate safely.
  • Publish versions for repeatable deploys and rollbacks.
  • Reuse jobs and blueprints instead of starting from scratch.

Deploy

  • Plan first: preview what will change and where it will run.
  • Roll out safely across workers (with targeted placement and scaling).
  • Rollback by deploying a previous published version.

Operate

  • See what’s running and where (workloads + fleet).
  • Track deployments end-to-end (events, progress, warnings).
  • Keep changes explicit: preview disruptive moves before applying them.
See workflows in detail →

Why these capabilities matter — Because your downstream tools shouldn’t dictate your data structure.

When pipelines live in vendor agents and per-tool configs, every change becomes fragile and expensive.

  • •Downstream tools no longer dictate your pipeline shape
  • •You avoid vendor lock-in while keeping the tools you already use
  • •You prevent cost explosions and surprise ingest bills
  • •You maintain one chain of custody for governance and audits

LyftData makes data predictable — operationally, economically, and in compliance posture.

See how teams put these capabilities to work

Explore Solutions →See Integrations →

Portability and stable references

Workflow node identities and blueprint aliases are treated as stable identifiers so placement/scaling references remain portable across edits and export/import loops.

See real workflows in action

Security, observability, and data teams already building on LyftData.

Check compatibility with your stack

Browse supported sources and destinations.

Understand the architecture

Walk through Server → Workers → Jobs in detail.