Data Analyst

Job Description:

About the role

We are building the analytics layer over our ERP app. You’ll own the analytics engineering core: modelling raw change-data into clean marts, instrumenting tests/docs, and shipping decision-grade dashboards that our users can reason over.

Model the warehouse

Design and implement stg_ , dim_ , fact_ , and mart_ layers in dbt for domains like HR, Finance/AR, Inventory, Manufacturing.

Define grain clearly; implement SCDs, incremental models, snapshots, seeds, exposures, and packages.

Create rigorous tests (unique, not null, relationships, custom macros) and data contracts.

ClickHouse performance

Optimize schemas, partitioning, primary keys/order by, materialized views; handle late-arriving data and CDC merge logic.

Own query performance and cost/perf trade-offs.

Dashboards that drive action

Build Superset datasets, charts, and curated dashboards; set conventions (naming, filters, time grains).

Contribute to/consume a chart registry exporter so charts/metrics are machine-readable by our LLM layer.

Governance & security

Implement RLS/RBAC patterns across Superset/ClickHouse; parameterize org/user scopes.

Keep dbt docs healthy; publish manifest/catalog; maintain docs_index.jsonl for downstream search

Ops & quality

Set up CI for dbt (build + tests on PR), freshness SLAs, run artifacts, observability.

Write excellent README/runbooks; coach stakeholders on metric definitions.

Must-have experience

2–6 years in Analytics Engineering / Data Warehousing.

Expert SQL and dbt (2+ years): modular modelling, incremental strategies, macros, exposures, docs, tests.

Strong dimensional modeling and clear articulation of grain.

Production experience with ClickHouse (or very strong columnar DB like BigQuery/Redshift/Snowflake plus willingness to go deep on ClickHouse fast).

Building BI artifacts in Superset (datasets, semantic fields, dashboards, filters).

Comfort with CDC concepts JSON extraction, merge patterns. Git, code review, and basic Docker workflows.

Nice to have

Python for analytics tooling (registry exporters, dbt utilities).

Experience with data quality frameworks (Elementary, Great Expectations, or custom dbt tests).

RLS patterns in ClickHouse/Superset; multi-tenant data modeling.

Understanding of LLM-aware analytics (documenting metrics/semantics so AI can reason over them).

Domain familiarity in any of: HR, AR/AP, Manufacturing/Inventory, Diamond/Jewellery, Textiles.

Tooling you’ll touch

Warehousing: ClickHouse Modeling:

dbt Core

BI: Apache Superset

Ops: GitHub Actions (CI for dbt), Docker

Lang: SQL, Python