Experience
Professional experience
Software Engineer II at Comcast: production distributed data platforms across hybrid AWS and on-prem environments — software engineering, platform reliability, automation, and operational responsibility.
Current role
Comcast
Software Engineer II · Current · 5+ years
5+ years writing and operating production software for distributed data platforms, with Java, Python, SQL, and shell development across a multi-cluster Trino / Starburst estate on AWS and on-prem.
Scale / operating environment
Where the work runs
- Trino / Starburst estate across two very different environments.
Multi-cluster
production fleet
Trino / Starburst estate across two very different environments.
- Fleet spanning cloud and on-prem.
Hybrid
AWS · on-prem
Fleet spanning cloud and on-prem.
- Large-scale distributed analytics over multiple sources.
Federated
data access
Large-scale distributed analytics over multiple sources.
- Production reliability and incident response.
On-call
incident response
Production reliability and incident response.
- Users keep querying the platform while releases progress.
Weekly
zero-downtime deployments
Users keep querying the platform while releases progress.
Engineering contributions
Specific work, by area
Query platform & engine work
- Java, Python, SQL, and shell development across the Trino codebase: proxy / integration components and query-performance engineering that keeps interactive analytics responsive.
- Client-side spooling: large result sets written as segments to object storage so clients download in parallel, offloading I/O from the coordinator.
- Pre-release query replay / validation: re-run stored queries against candidate builds to catch regressions before rollout.
Data access & security
- Snowflake on-behalf-of authentication through Trino, so users reach data with their own identities.
- Privacera-based access control and OAuth / identity integration, enforcing authorization at the data layer.
- Iceberg permissions and access-control work; security remediation across the toolchain (Snyk, CrowdStrike).
Infrastructure & delivery
- Terraform + Ansible infrastructure-as-code kept declarative and reviewable across AWS and on-prem.
- GitHub Actions CI/CD with pipeline-gated, reversible rollouts; Kubernetes / Helm for platform workloads.
- S3-related platform migration work that keeps the estate portable and operational.
Client & ecosystem integration
- Databricks & Unity Catalog — hands-on troubleshooting and integration work: validating communication between Databricks and the distributed query platform, including connectivity, authentication / access, and compatibility with the Trino / Starburst platform.
- Heterogeneous consumers — supported Snowflake, Spark, Tableau, Python clients / connectors, and SSIS, spanning authentication, connectivity, query behavior, and platform compatibility.
Reliability & impact
What changed
Mean time to recovery
~3h → ~1h
through better observability and automation.
Production incident response
- Fleet-scale Prometheus / Grafana dashboards and alerting that support fast diagnosis across clusters.
- Runbooks and automation that remove repeat failure classes instead of paging through them.
Agentic AI / security · professional
Agentic security & compliance
Professional agentic AI / security work, separate from the personal lab: an AI-driven security & compliance platform whose agentic workflows analyze repositories, assess risk, and surface structured remediation guidance.
Personal lab
Heterogeneous AI inference lab
Outside of professional work, I run a heterogeneous multi-GPU lab where the same systems discipline — memory budgets, engine selection, quantization, observability — is applied to local LLM inference and model serving.
CUDA · SGLang · vLLM · llama.cpp · Quantization (NVFP4 · FP8 · GGUF) · Speculative decoding · KV-cache & long-context engineering · Inference benchmarking
Education
Drexel University
Master of Science in Software Engineering
- Graduate Certificate in Software Architecture
- Graduate Certificate in Artificial Intelligence & Machine Learning
- Graduate Certificate in Software Security
Stack
Stack, organized by context
Production / CoreJava · Python · SQL · Bash / Shell · Trino / Starburst · Distributed query processing · Query performance / optimization · AWS · Linux · Terraform · Ansible · GitHub Actions · Prometheus · Grafana
Platform / InfrastructureKubernetes / Helm (workload hosting) · Docker · CI/CD pipelines · Infrastructure as code · OAuth / access control · Security & access tooling (Snyk, CrowdStrike)