- Architected and delivered 100+ APIs in .NET 9 supporting multi-source data ingestion, processing, and analytics for vendor master data quality.
- Designed a shared NuGet package (Smartvm.Common.Components) for reusable components, versioned via Azure Artifacts to accelerate delivery across services.
- Built high-performance APIs with Swagger/OpenAPI, gRPC for internal calls, and caching (Memcached, .NET Backplane); improved response times and reduced database load.
- Deployed to Azure Kubernetes Service with Helm, containerized builds, Key Vault–backed secrets, and horizontal scaling for demand spikes.
- Implemented OpenTelemetry with Application Insights and Grafana dashboards for proactive monitoring and faster troubleshooting.
- Integrated RabbitMQ and Azure Service Bus for event-driven messaging and reliable asynchronous processing.
SmartVM Subscriptions Service — API Quotas & Access Control
- Designed and implemented a dedicated API management microservice controlling access, quotas, and product-based API grouping.
- Managed per-client API quotas, restrictions, and expiration policies; enforced via middleware before request execution.
- Built background services for quota monitoring and automated email alerts for key expiration and usage thresholds.
- Exposed as a Kubernetes ClusterIP service for internal consumption by API platform services.
SmartVM Data Processing & Ingestion Pipelines
- Developed .NET 9 Worker Services for ingesting data from SQL, REST APIs, and web scraping pipelines using Playwright.
- Scheduled distributed ETL jobs via Hangfire with retries; deployed via Kubernetes with min/max replicas and readiness/liveness probes.
- Created transformations to centralize vendor data into SQL Server for downstream analytics; integrated internal SSIS pipelines for legacy migrations.
SQL / Azure Blob / Snowflake Integration
- Implemented .NET-based data pipeline to stage SQL data in Azure Blob (Parquet/CSV) and load into Snowflake.
- Automated COPY INTO and Snowpipe ingestion; used merge/upsert logic for incremental loads and late-arriving data.
- Reduced transfer time with partitioned extracts, compression, and multi-threaded Blob uploads.
Snowflake Processing — Vendor Matching & Golden Records
- Developed Snowflake stored procedures, Java UDFs, and JavaScript transformations based on analyst-defined rules.
- Implemented matching and confidence scoring logic; generated golden vendor records to improve master data quality.
Snowflake / SQL Reverse Data Flow
- Implemented bulk unload from Snowflake to Blob, then loaded into SQL Server via BULK INSERT/PolyBase for reporting needs.
SmartVM Web Application (Full-Stack)
- Developed a .NET 9 MVC + AngularJS application for managing and reviewing vendor data runs.
- Enabled user uploads, real-time data quality metrics, and integration with backend APIs for enrichment.
- Supported exporting processed data to files or downstream internal products.