Agentic Analytics Products Developer

ebayinc

Bengaluru, India 3 Years Exp Posted 34d ago

What you’ll do

Build data intelligence services: design FastAPI endpoints that expose metrics, diagnostics, and insights as composable analytical functions.
Operationalize data access: connect to warehouses and lakes, generate efficient SQL or DuckDB queries, and design Parquet pipelines that surface clean, ready-to-analyze data.
Enable analytical reasoning: support retrieval-augmented generation and grounding so that every AI-produced statement can be backed by verifiable numbers, tables, or queries.
Improve insight reliability: define schemas, metadata, and lineage that make it clear where data came from, how it was transformed, and what assumptions were applied.
Measure analytical performance: instrument pipelines with metrics for latency, cost, freshness, accuracy, and coverage; define thresholds that balance insight quality with compute efficiency.
Partner across disciplines: work with data scientists, AI developers and Analytics Product Managers to implement analytics products.
Ensure compliance and governance: implement privacy-aware joins, role-based access, and redaction of PII while maintaining analytical reproducibility.

3+ years building analytics pipelines, APIs, backend-for-frontend and insight-serving systems.
Deep understanding of SQL, data modeling, query optimization, and columnar formats (Parquet, DuckDB).
Experience turning analytical logic (diagnostics, KPIs, A/B metrics) into reusable services or models.
Familiarity with LLM/agent integrations for analytics (structured tool calls, grounded outputs).
Strong observability mindset — tracking data freshness, success rates, and cost per insight.
Commitment to data accuracy, reproducibility, and explainability.

Background in applied AI, data science Analytics, BI, or experimentation platforms.
Exposure to LLM prompt optimization and human evaluation.
- Experience in software deployment (e.g. Kubernetes).