Staff Data Engineer

Ovii's Interpretation of the Role

Staff Data Engineer leading the design and implementation of Sarvam's Data & Analytics Platform. Owns end‑to‑end ingestion, CDC pipelines, and multi‑tenant query APIs that serve product, BI, and billing workloads.

Role Snapshot

• Technical owner of data platform
• Design & build high‑throughput ingestion APIs
• Implement CDC pipelines with Debezium & Kafka
• Develop multi‑tenant query service
• Shape engineering culture and hiring standards

Must-Have Requirements

• ClickHouse
• Golang
• Kafka
• Debezium
• CDC pipelines
• backend engineering
• analytics platform design
• production data systems
• On-site work in Bengaluru

Nice-to-Have Signals

• Spark
• Airflow
• Multi‑tenant SaaS patterns
• BI/dashboard tools (Superset, Metabase, Cube)
• Billing / metering data exposure
• batch processing (Spark, Airflow)
• multi‑tenant SaaS platforms
• BI/dashboard integration

Work Setup

• Location: Bengaluru, India
• Work mode: ONSITE
• Remote scope: UNSPECIFIED
• Employment type: Full-Time

Not Specified in JD

• Visa sponsorship
• Salary range
• Remote eligibility
• Education requirement
• Certifications
• Relocation
• Notice period
• Travel
• Security clearance
• Coding test
• Portfolio
• GitHub
• Writing sample
• Cover letter

What You'll Likely Work On

• Create a Go‑based event ingestion API that auto‑evolves schemas and writes to ClickHouse.
• Build and operate Debezium‑driven CDC pipelines streaming from Postgres/MySQL into ClickHouse.
• Design a tenant‑aware query API with RBAC, query budgets, and isolation guarantees.
• Implement a no‑code dashboard layer for customers and internal teams.
• Engineer ClickHouse cluster (sharding, replication, materialized views) and manage capacity, observability, and SLOs.
• Define design‑review processes, RFCs, and testing standards for the platform team.

Good Fit If You Have

• Enjoys owning foundational infrastructure rather than feature work.
• Balances code writing with thorough design documentation.
• Focuses on cost‑per‑query and operational efficiency.