AI-Powered Splunk Traces Migration to datadog
Budget: $30 – $250 CAD
I need to move all existing trace data out of Splunk and into Datadog, and I want the process to lean on AI wherever it adds real value—from automated field discovery to error-pattern classification. The scope is limited to traces only; logs and metrics will remain in place for now. During the migration, every trace must be mapped by Service name so that dashboards, monitors, and any downstream analytics in Datadog continue to resolve correctly without manual retagging.
Here’s how I picture the engagement:
• You design and build a repeatable pipeline—Python, Go, or a low-latency ETL tool of your choice is fine—capable of extracting historical traces from Splunk, transforming them to a Datadog-ready schema (OpenTelemetry or Datadog APM format), and then importing them through the Datadog API.
• Artificial-intelligence assistance is expected. I’m interested in practical AI touches such as anomaly detection on sample data to validate completeness, or an LLM-backed mapping assistant that flags ambiguous service names. Feel free to propose creative approaches, but keep the outcome measurable and auditable.
• Final delivery includes a runnable script or container, minimal configuration docs, and a verification report that proves trace counts match across both platforms when grouped by Service name.
If you’ve tackled similar Splunk → Datadog migrations—or any trace migrations involving OpenTelemetry—let me know. I’m ready to get started as soon as the plan looks solid.
Here’s how I picture the engagement:
• You design and build a repeatable pipeline—Python, Go, or a low-latency ETL tool of your choice is fine—capable of extracting historical traces from Splunk, transforming them to a Datadog-ready schema (OpenTelemetry or Datadog APM format), and then importing them through the Datadog API.
• Artificial-intelligence assistance is expected. I’m interested in practical AI touches such as anomaly detection on sample data to validate completeness, or an LLM-backed mapping assistant that flags ambiguous service names. Feel free to propose creative approaches, but keep the outcome measurable and auditable.
• Final delivery includes a runnable script or container, minimal configuration docs, and a verification report that proves trace counts match across both platforms when grouped by Service name.
If you’ve tackled similar Splunk → Datadog migrations—or any trace migrations involving OpenTelemetry—let me know. I’m ready to get started as soon as the plan looks solid.