Remote opportunity at
DeepgramSenior Analytics Engineer - Decision Intelligence
Deepgram sells real-time speech-to-text and text-to-speech APIs, along with tools for building voice agents, available through the cloud or as self-hosted software. Its developer base is…
Career Tools
About This Role
Deepgram sells real-time speech-to-text and text-to-speech APIs, along with tools for building voice agents, available through the cloud or as self-hosted software. Its developer base is said to exceed 200,000, with 1,300-plus organizations building voice products on its platform. This Senior Analytics Engineer role supports the Decision Intelligence effort, which is creating a governed set of metrics that humans and AI agents can rely on without redoing the math. The…
Job Description
Deepgram sells real-time speech-to-text and text-to-speech APIs, along with tools for building voice agents, available through the cloud or as self-hosted software. Its developer base is said to exceed 200,000, with 1,300-plus organizations building voice products on its platform. This Senior Analytics Engineer role supports the Decision Intelligence effort, which is creating a governed set of metrics that humans and AI agents can rely on without redoing the math.
The engineer works one layer above the warehouse. Ingestion and orchestration are handled by the data platform team, so this job focuses on modeling. It involves building conformed gold-layer models for customers, accounts, projects, products and contracts, and reconciling identities across Salesforce, the company's console, Stripe and its billing system, none of which share an authoritative key. The engineer also owns the semantic layer, defining each metric once so dashboards, notebooks and agents agree, and versions those definitions over time.
Additional work includes building the tools, permissions and interfaces that let analyst agents query data without guessing at joins, and making correctness testable through assertions on grain, uniqueness and cross-system reconciliation. The engineer also grades agent answers against known-correct results, tracking accuracy, refusals and confidently wrong responses as separate numbers. Questions from sales, product and finance get answered, and every answer should yield a model that shortens the following request.
The role is a senior technical individual contributor job on a team that is still being formed. The listing is marked remote. The posting does not state a salary range or describe benefits.
Responsibilities
- Build gold-layer models covering customers, accounts, projects, products and contracts.
- Own the semantic layer, define each metric once, and version or retire definitions as needed.
- Build the agent-facing tools, permissions and interfaces that keep analyst agents from inventing joins.
- Add assertions for grain, uniqueness and cross-system reconciliation, with a named owner for each failure.
- Grade agent output against known-correct results and track accuracy, refusals and confidently wrong answers.
- Answer questions from sales, product and finance, and leave behind reusable models.
Requirements
- AI-native working habits, applied daily and in depth, with a clear example of recent work built with AI.
- Production-grade SQL and Python, including scheduled code that has broken and been caught by the engineer's own monitoring.
- Ownership of a semantic or metrics layer using tools such as dbt, Cube or LookML, including a change to a definition people already relied on.
- Modeling judgment about grain and keys before writing a query, and the discipline to separate true data bugs from open business questions.
- Comfort building on unfinished infrastructure with no existing ticket queue.
- Clear writing for definitions and data contracts that others must trust.
Qualifications
- Experience with usage-based billing, where committed, consumed, invoiced and recognized revenue are distinct figures.
- Experience with real-time billing or consumption systems.
- Experience with lakehouse architectures such as Iceberg, Athena or Trino.
- Public code for open-source data or AI efforts.
Core Skills
Frequently Asked Questions
Answers are based only on the employer’s listing; where it doesn’t say, neither do we.
Is this role remote?
The listing is marked remote. The posting does not name a specific office location.
Is a salary listed?
No. The posting does not state a salary range or describe benefits.
Does the role expect AI use?
Yes. The posting says every team member is expected to use AI tools actively, and that AI use is measured as part of performance.
How do I apply?
The posting text does not describe the application steps. It notes that legitimate Deepgram recruiting messages come from an @deepgram.com address.
Sample Interview Questions
AI-generated questions tailored to this specific role — a preview of the full practice set.