Core0 jobs and careers

About Core0

AI systems fail differently from traditional software. A rogue agent makes an innocent-looking database query that chokes the DB, congests the network, and triggers a cascade of downstream failures — none of which appear in the agent's own logs. Meanwhile, your telemetry bill keeps climbing while your team still can't find root cause fast enough. FFWD is the AIOps platform built for this environment. We give engineering teams four interlocking capabilities on one stack: High-performance telemetry pipeline — ingest, filter, enrich, and route 500,000+ events per second, 20+ TB per day, at a fraction of Splunk or Datadog cost. Built in Rust. No vendor lock-in. Full-stack anomaly detection — our Holistic Log Model applies six distinct ML model categories simultaneously across logs, metrics, traces, GPU telemetry, and container events. No rules to write. No topology map required. Operational from day one. Cross-domain correlation — when anomalies appear in unrelated systems at the same time, FFWD's marker correlation engine finds the statistical link. Root cause hidden in a GPU thermal trend, a sequence anomaly in a Python service, and a network rate spike? FFWD connects them. Agent Assurance — the only platform that runs a pre-task risk score on autonomous AI agents before they execute consequential actions, and a confidence score after. A Go/No-Go gate for every agent, grounded in live full-stack anomaly intelligence — not static rules. These aren't four products. The pipeline feeds the anomaly model. The anomaly model powers the risk score. The risk score is only trustworthy because the data underneath it is complete. Used by engineering teams who are deploying AI agents at scale and need to trust what they're doing. MCP-native. eBPF zero-instrumentation. Air-gap and private cloud ready.