Sagnik Ghosh
Data Analyst with 2+ years of experience delivering data-driven insights for enterprise clients. Expert in SQL, Power BI, Python, PySpark, and Azure/Microsoft Fabric data platforms for building scalable ETL/ELT pipelines, analyzing large-scale datasets, and building dashboards.
Experienced in data pipeline orchestration, data quality and audit frameworks, stakeholder management, and reporting automation to support business decision-making.
Experience
Data Analyst
Polestar Analytics
Bengaluru
Objective: Analyzed and processed 5M+ rows of marketing and sales data to track campaign performance, revenue trends, and deliver actionable KPI insights to business stakeholders.
Technologies: SQL, Power BI, DAX, Python (Pandas, NumPy), Azure Databricks, ADLS Gen2, Azure Data Factory, Microsoft Fabric
Designed 10+ Power BI dashboards and semantic models, improving report accuracy by 25% and stakeholder adoption by 30%.
Built reusable SQL queries and DAX measures to standardize KPI tracking across teams.
Performed data cleaning, transformation, and validation, reducing reporting errors by 40%.
Defined 15+ KPIs in collaboration with business stakeholders and delivered actionable insights.
Optimized data pipelines, cutting reporting turnaround time by 50%.
Performed Exploratory Data Analysis (EDA) to identify performance drivers and optimization opportunities.
Education
Jadavpur University
B.E. in Engineering
CGPA: 8.0/10
Skills
PERSONAL PROJECTS
End-to-End Sales Data Pipeline
Microsoft Fabric, PySpark, Delta Lake, SQL Server, Medallion Architecture, Data Pipeline Orchestration
Architected and orchestrated an end-to-end Microsoft Fabric data pipeline (21 activities) using a medallion (Bronze-Silver-Warehouse) lakehouse architecture, with conditional branching, idempotency checks, and automated file/schema validation.
Built PySpark ETL/ELT transformations for data cleansing, deduplication (window functions), type standardization, and business-rule-driven enrichment, writing curated datasets to Delta Lake tables.
Developed a custom audit and error-logging framework (SQL Server stored procedures and audit tables) to track pipeline run status, row counts, and stage-level failures for full data lineage and reconciliation.
Implemented idempotent, audit-driven warehouse loads with post-load data validation, parameterizing the pipeline for seamless multi-environment (DEV/TEST/PROD) deployment.
E-Commerce Data Pipeline
Databricks, Lakeflow Declarative Pipelines, GitHub, DAB
Engineered an end-to-end Lakeflow declarative pipeline on Databricks to process 100K+ e-commerce records from the Olist Brazilian dataset across orders, payments, sellers, and customer reviews, implementing data quality expectations and constraints through Spark Declarative Pipelines.
Managed infrastructure as code using Databricks Asset Bundles (DAB) with YAML-based job configuration, maintaining a fully version-controlled deployment workflow via GitHub CI/CD.