Ops
Cost Anomaly Dashboard
Real-time spend monitoring with automated Slack alerts for cloud resource over-allocation.
Real-world data engineering and platform reliability projects.
Built automated evaluation and monitoring workflows for tracking model hallucinations and prompt performance.
Implemented batch/stream feature generation with lineage and reproducible backfills.
Event-driven processing with quality checks and low-latency serving for analytics consumers.
Built trusted transformation models and orchestration for reliable cross-functional reporting.
Terraform modules for standardized K8s cluster lifecycle management across environments.
End-to-end retrieval-augmented generation pipeline using Pinecone, OpenAI, and LangChain for domain-specific search.
Real-time spend monitoring with automated Slack alerts for cloud resource over-allocation.
Automated schema drift alerts and rule-based validation for upstream data contracts.