# One data layer for security, observability, and AI agents.

CtrlB is one lake of open Parquet in object storage you own, and four stages over it. Collect what your fleet emits, search all of it in seconds, let the system surface what changed, and give your agents the same access your engineers have.

## Metadata

- **Canonical HTML**: https://ctrlb.ai/platform

## At a glance

| Figure | What it refers to |
|--------|-------------------|
| PB scale | On object storage you own |
| 4 components | One data layer |
| Open Parquet | No proprietary index |
| SaaS · BYOB · BYOC | Three deployment modes |

## Core components

| Component | Role | Page |
|-----------|------|------|
| Collect & Control | Configure, update, and govern every collector and agent from one place. | https://ctrlb.ai/platform/collect |
| Search & Investigate | Full-text search, SQL, metrics, and traces on object storage. | https://ctrlb.ai/platform/search |
| Analyze & Correlate | Patterns, anomalies, and correlations across your telemetry. | https://ctrlb.ai/platform/analyze |
| Agent Access | Give any agent fast access to complete operational history. | https://ctrlb.ai/platform/agent-access |

### Collect & Control

_One control plane for every collector and agent._

Configure, update, and manage your collector and agent fleets from one place. Roll out changes across clusters and clouds without touching every host individually.

- Fleet-wide configuration for OpenTelemetry collectors and agents.
- Staged rollouts across clusters and clouds, with the option to roll back.
- Health and version visibility for every collector you run.
- Governs both the collectors feeding the platform and the agents querying it.

- **Dedicated page**: https://ctrlb.ai/platform/collect

### Search & Investigate

_One engine for search, SQL, metrics, and traces._

The Data Engine stores, indexes, searches, and analyzes everything you send. Telemetry lands as open Parquet in object storage you own, and compact inverted indexes are written beside it as data arrives — so a needle query reads an index and a handful of row groups rather than scanning the lake.

- Full-text search over raw log bodies, with no fields to declare in advance.
- Standard SQL for aggregation, joins, and windows across every signal.
- Metrics stored as columns rather than series, so cardinality has no ceiling.
- Stateless query nodes that fan out on demand and scale to zero when idle.
- Schema-less ingestion: new fields index the moment they appear.
- First place on the combined ClickBench result, untuned.

- **Dedicated page**: https://ctrlb.ai/platform/search

### Analyze & Correlate

_Patterns, anomalies, and correlations across your telemetry._

AI Insights runs over the whole lake rather than a sampled recent window. It collapses repetitive log lines into the handful of patterns actually behind them, flags where behaviour departs from its own history, and connects movements across logs, metrics, and traces that would otherwise be three separate investigations.

- Repetitive logs compress into patterns a human or an agent can reason over.
- Anomalies are measured against full history, not the last fifteen days.
- Correlations span signals, because all three live in one data layer.
- Findings are queries you can open, inspect, and adjust — not opaque scores.

- **Dedicated page**: https://ctrlb.ai/platform/analyze

### Agent Access

_One MCP endpoint over the whole lake._

CtrlB does not sell you an agent. It exposes a Model Context Protocol interface over the same data layer the UI queries, so the agent you already trust reads complete operational history directly — no dashboard scraping and no export pipeline to keep in sync.

- Bring your own agent; the interface is open rather than a bundled assistant.
- Sub-second responses, so an agent can ask twenty questions instead of three.
- Agents and engineers read the same evidence from the same files.
- Sustained query volume is a compute decision, not a licensing tier.

- **Dedicated page**: https://ctrlb.ai/platform/agent-access

## Deployment

The same platform runs in three shapes. See the dedicated page: https://ctrlb.ai/platform/deployment

### SaaS

We run it. You point telemetry at it.

- Fastest path to a working lake — no infrastructure to stand up.
- Storage and compute operated by CtrlB in the region you choose.
- Right for teams who want the query layer without the operations.

### BYOB — bring your own bucket

Your object storage. Our query layer.

- O2C reads Parquet from your bucket and ingests into CtrlB’s query layer.
- Indexes are written back to your bucket beside the raw data — both stay under your control.
- Scoped IAM only; retention, lifecycle, and object-lock policies remain yours.

### BYOC — bring your own cloud

Everything runs inside your account.

- Query and ingest components run in your cloud, next to your data.
- Choose fully private (UI + auth in your account) or managed UI (Flow and SSO on our cloud).
- Telemetry never crosses a network boundary you do not control.

## Security and enterprise controls

- **Encryption everywhere**: Encryption at rest and in transit, with hierarchical keys and private networking between components.
- **Data residency**: Choose the region, or run entirely inside your own account under BYOB or BYOC so telemetry never leaves it.
- **Immutable storage**: Write-once Parquet with object-level locking and no compaction rewrites, so the audit trail stays intact.
- **One copy, two jobs**: Compliance retention and live investigation read the same objects, removing an entire class of reconciliation work.
- **Reliability**: Uptime SLAs and mission-critical support for scale-out deployments.
- **Compliance posture**: Industry-standard security practices and compliance with privacy regulations; specifics available on request.

## Try it

- **Live playground**: https://flow.ctrlb.ai/playground
- **Markdown mirror**: Fetch `/platform.md` for this page as plain Markdown