Analytics Engineer (AI)

Location

Location: USA | Remote

Company

Deepgram is a leading Voice AI platform providing real-time speech-to-text and text-to-speech APIs, serving over 1,300 organizations with advanced voice-native foundation models.

What you will do

• Own the metric registry by establishing canonical metric definitions and resolving conflicts.
• Implement enforceable definitions in semantic layers and data catalogs, retiring outdated versions.
• Audit and reduce the reporting estate, establishing clear ownership.
• Build data quality checks including freshness, uniqueness, and referential integrity with assigned owners.
• Verify AI agent outputs against ground truth and maintain an inventory of data-accessing agents.
• Enable self-serve access to governed and trustworthy data for users and AI tools.

Requirements

• Must have 5+ years in analytics or related fields.
• Strong SQL skills including reverse-engineering undocumented logic.
• Direct ownership experience of semantic or metrics layers (dbt, Cube, LookML, or equivalent).
• Proven ability to resolve conflicting metric definitions and communicate clearly in writing.
• Experience evaluating LLM or AI agent outputs against ground truth.
• Experience with data catalogs, lineage tooling, and lakehouse architectures (Iceberg, Athena, Trino).

Nice to have

• Exposure to audit readiness, SOX, or financial controls environments.
• Experience with consumption or usage-based business models.
• Experience joining early-stage functions before processes existed.