I design systems that move data at scale: pipelines that do not miss SLAs, streams that do not drop events, and infrastructure that deploys itself.
Currently, I work at Cisco on network observability. Previously, I helped build the data layer for a national-scale payments platform.
Pipeline Orchestration 路 Stream Processing 路 Batch ETL at Scale 路 Event-Driven Architecture
Distributed Systems 路 Data Lake Design 路 Multi-Cloud Deployments 路 Infrastructure as Code
Failure Recovery 路 SLA Monitoring 路 CI/CD Automation 路 Open Source
Airflow 路 Spark 路 PySpark 路 Flink 路 Kafka 路 Hudi 路 Trino
Python 路 Scala 路 Kubernetes 路 Terraform 路 Helm
AWS 路 GCP 路 Azure
I鈥檓 a data engineer with 3 years of experience across payments infrastructure, open source data platforms, and network observability.
My work sits mostly in the orchestration, transformation, and deployment layers, using tools like Airflow, Spark, Flink, Kafka, Kubernetes, Helm, and Terraform.
In my first role, I owned Spark pipelines for a high-volume payments platform. I learned what it takes to keep large batch workloads reliable, optimized, and on schedule, while contributing to significant reductions in compute usage.
I also worked as a backend contributor to an open source data platform, shipping 100+ pull requests across:
- Stream processing jobs
- Kafka connectors
- APIs
- Multi-cloud deployment tooling
I鈥檓 currently building network telemetry pipelines at a large tech company, working across the full stack from ingestion to deployment.
-
Airflow orchestration
Designing DAGs that handle real-world failures from the start. -
Spark optimization
Rewriting slow or resource-heavy workloads to improve performance and reduce compute usage. -
Kafka-based stream processing
Building high-throughput event pipelines that are reliable under load. -
Data lake tooling
Working with Hudi and Trino for scalable data lake architectures. -
Cloud infrastructure & CI/CD
Using Kubernetes, Helm, Terraform, and CI/CD pipelines to make deployments repeatable across AWS, GCP, and Azure.
Debugging is something I鈥檝e gotten good at out of necessity. When a pipeline breaks in production, I鈥檓 usually the one tracing it back to the source.
