Vision
Willy Lulciuc
We're building an AI Data Reliability Engineer (DRE) that uses autonomous agents to assist in data incidence resolution with full context of your ML/data stack.
- Correlates metrics, logs and changes instantly. Zero context switch.
- Identifies root cause hypothesis with suggested fix
- Maps your data infrastructure in real-time via a versioned, run-level lineage graph
The "data context" gap in data engineering
For years, data engineering followed a predictable, linear path. We built pipelines, defined a run schedule 0 * * * * and monitored health through dashboards with siloed context.
When a pipeline failed, a human paged through logs and dashboards to visually correlate a code change to a downstream spike in errors. It was a manual, contextless loop of debugging: one 2AM alert, one engineer and endless context switching between dashboards. Investigation moved at the speed of human intuition.
Beyond the dashboard: The AI era of evidence based investigations
As data infrastructure grows in complexity, the "data context" gap widens, fast. A human can't keep the entire data dependency graph in their head. At 2AM, the investigation becomes unmanageable as existing observability tools fail to close the root-cause loop.
This is what oleander will do. Oleander is defining the era of AI-native observability by providing an independent telemetry context layer for ML/data infrastructure, an interface where humans and agents can learn and debug together.