Cardinal provides an observability data lake that stores full‑fidelity logs, metrics, traces, and events directly in a customer’s object storage using OpenTelemetry‑native ingestion. Its Lakerunner indexer creates zero‑egress, high‑cardinality indexes, while AI agents in a secure runtime enable natural‑language investigations and deterministic root‑cause analysis with evidence‑backed answers, all without vendor lock‑in or query limits.
Funding
$3.5M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Modern observability stacks require sending logs, metrics, traces, and events to third‑party SaaS platforms, incurring data egress costs, vendor lock‑in, and query limits that hinder deep root‑cause analysis across large, high‑cardinality datasets.
Solution
Cardinal provides an observability data lake that stores full‑fidelity telemetry directly in a customer’s own object storage. An OpenTelemetry‑native exporter streams data to the lake, where the Lakerunner indexer creates searchable indexes in place without data egress. AI agents run inside a secure runtime, connecting to the indexed data to perform natural‑language investigations, deterministic troubleshooting, and evidence‑backed answers. The platform automatically correlates change events, clusters errors, and computes statistical baselines, enabling agents to surface root causes across months of data with sub‑second latency while keeping all data within the user’s VPC.
Target Audience
Primary customers are SRE and DevOps teams at mid‑size to large enterprises that manage high‑volume, high‑cardinality observability data and require on‑premise, AI‑driven root‑cause analysis.
Features
- OpenTelemetry‑compatible ingestion that writes logs, metrics, traces, and events directly to object storage in Parquet format
- Lakerunner indexer creates full‑fidelity, zero‑egress indexes in place, supporting ultra‑high cardinality queries without query rate limits
- Agent Runtime with pre‑built troubleshooting skills (e.g., metric spike decomposition, change‑event correlation, fingerprint‑based error clustering)
- Composite skills that execute hundreds of queries in seconds, returning structured, impact‑ranked attributions with evidence
- Built‑in knowledge graph that captures entities, relationships, and decision traces to improve future investigations
- Support for any cloud‑hosted LLM endpoint (Claude, ChatGPT, Vertex, Azure OpenAI) via per‑token pricing, ensuring data never leaves the VPC
- Unlimited retention and query volume under a perpetual license, with managed deployment assistance and auto‑scaling compute