Mantri Srinivas

Pricing & Inventory Analytics — VINtelligence data ingestion, Databricks-based pricing intelligence, and Synapse reporting for the auto-parts platform.

Python, PySpark & Databricks

• Wrote standalone Python scripts for data validation, secret retrieval via Azure Key Vault SDK, and utility automation tasks outside the Spark context — directly addressing Python-as-scripting requirement.

• Designed a Customer 360 data model on Azure Databricks integrating transactional, clickstream, and support-ticket data into unified Delta tables with SCD Type 2 history for customer profile changes.

• Built a near-real-time personalisation feed using Azure Event Hubs and Spark Structured Streaming, scoring customer sessions for cross-sell recommendations within seconds of an event.

• Used Auto Loader with schema evolution for incremental file ingestion; applied AQE, broadcast joins, and Z-Order/OPTIMIZE on high-traffic Gold tables to maintain dashboard query performance as data volumes grew.

Delta Live Tables & Unity Catalog

• Added DLT expectations to validate order totals, currency codes, and duplicate transaction IDs, cutting weekly reconciliation tickets from ~12 to a handful — concrete DLT data-quality gating.

• Worked with the Data Governance team to define Unity Catalog access policies (schema-level access control, PII column masking) on owned tables and registered pipelines for lineage tracking.

Databricks Asset Bundles (DAB) & CI/CD

• Collaborated with the DevOps team to deploy Databricks notebooks and Delta Live Tables using Databricks Asset Bundles (DAB) on the Azure DevOps release pipeline — Dev → UAT → Prod promotion with parameterised job configs.

• Maintained bundle YAML manifests for cluster policies, job configs, and pipeline definitions, enabling version-controlled, repeatable Databricks deployments.

Cost & Performance Optimisation

• Right-sized Databricks cluster policies (auto-scaling limits, spot-instance mix, job vs. all-purpose clusters) and adopted Photon runtime for SQL-heavy notebooks, meaningfully reducing compute spend.

• Investigated failed Databricks job runs using cluster event logs and Spark UI; resolved SLA-impacting incidents as part of the on-call rotation.

Python & PySpark on Databricks

• Wrote Python utility scripts for file-arrival monitoring, secret management via Azure Key Vault SDK, and automated post-load validation checks outside the Spark execution context.

• Built PySpark pipelines on Azure Databricks to compute dynamic pricing recommendations from VINtelligence transaction and competitor-pricing data, using Delta Lake for versioned pricing history.

• Scheduled nightly Databricks Workflows jobs to refresh inventory-forecasting datasets for the data-science team's Azure ML model.

Azure Data Factory

• Designed parameterised ADF pipelines using Linked Services, Datasets, Copy, Lookup, Conditional Split, Filter, Join, and Union activities.

• Automated ingestion of VINtelligence flat files from FTP into ADLS Gen2 with archival workflows for compliance; replaced legacy SSIS ETL with ADF + Databricks notebooks.

Azure SQL / Synapse & Security

• Developed and optimised T-SQL stored procedures, functions, and views for auto-parts groupings, pricing updates, and inventory lookups; implemented SCD Type 1 & 2 and PolyBase external tables.

• Created a Synapse serverless SQL pool layer over curated Delta tables for ad-hoc analyst queries; implemented row-level security in shared Power BI datasets for regional managers/dealers.

• Used Azure Key Vault for Linked Service secrets and configured RBAC on Azure SQL and Synapse for least-privilege, read-only dealer-facing access.

• Managed production deployments via Git (Azure Repos), Redgate script generation, and CI/CD pipelines; triaged failed ADF runs and SQL Agent failures with hotfix and rollback procedures.

Athena Health — Patient Registration, Appointment, and Billing modules for healthcare data integration.

• Performed data profiling and validation for Athena Practice and Athena Flow healthcare applications.

• Designed SSIS ETL packages for extraction, validation, and loading against Athena modules (Patient Registration, Account, Appointment).

• Implemented ADF pipelines for medical billing data extraction from Azure Blob Storage; managed deployments across test, UAT, and production environments.

• Ingested satellite imagery, cadastral data (unstructured), and land transaction logs into ADLS Gen2 using ADF pipelines; processed TB-scale geospatial datasets using PySpark in Databricks.

• Applied PySpark transformations to cleanse and classify geospatial datasets; enabled detection of illegal land encroachments and land-use optimisation through advanced spatial analytics.

• Implemented SCD Type 2 for historical land ownership and zoning records; implemented CDC for DML audit history; designed SQL stored procedures and migration scripts for core land-management logic.

• Delivered Power BI reports with geospatial visualisations and interactive dashboards for urban-development policy-making.

• Set up Git feature branches, Azure Repos PRs, code reviews, and CI/CD best practices; mentored and task-allocated to junior developers.

Workers Comp & Smart Audit — integrated claims management for workers compensation and auto liability.

• Analysed requirements; created and altered SQL Server objects (tables, stored procedures) with unit testing via TFS source control.

• Designed ETL-SSIS packages for Fee Schedule module (sdup, suds, pswi formats) with incremental loading and performance tuning; developed ADF pipelines with SCD Type 1 & 2.

• Collaborated with business analysts to translate requirements into technical ETL specifications; mentored junior team members.

• Developed interactive Power BI reports and dashboards to support executive decision-making, sales analysis, finance reporting, and operational KPIs.
• Designed and published Power BI reports using Power BI Desktop, leveraging advanced data modeling techniques including star schema and relationship management.
• Created complex DAX measures, calculated columns, and calculated tables for YTD, MTD, YoY growth, rolling averages, running totals, ranking, and KPI calculations.
• Built reusable semantic models and optimized report performance by implementing efficient DAX formulas, reducing report refresh time by up to 35%.
• Developed drill-through reports, bookmarks, tooltips, slicers, and dynamic visualizations to enhance user experience.
• Implemented Row-Level Security (RLS) to restrict data visibility for regional managers, dealers, and business users based on organizational hierarchy.

Samaritan Health Plans — reporting across network hospitals, doctors, and specialists.

• Designed and maintained Crystal Reports with complex multi-database SQL queries; optimised report performance through SQL tuning and best-practice report design.

SDZ — Australian Government governance workflow and data dictionary management.

• Implemented Telerik Reporting with data connectivity, binding, parameterisation, and multi-format rendering (HTML, PDF, Excel, Word, CSV).

• Managed Azure Blob Storage; built ADF pipelines for JSON ingestion and used OPENJSON to convert JSON into structured SQL Server format.

Telecom domain — Indus Towers & ATC passive infrastructure WFM data (BTS, MW, RTT, GBT).

• Developed comprehensive SSIS ETL processes for CSV-based WFM data: Extract, Transform (Derived Column, Conditional Split, Aggregate, Lookup), and Load.

• Wrote complex T-SQL queries and stored procedures; optimised database performance through indexing, bulk inserts, and query tuning; scheduled SSIS packages via SQL Server Agent.