Senior Software Engineer, AI Platform at Google

Owned end-to-end AI product features for the Gemini Multimodal Document Intelligence Platform, supporting regulated industries including Finance, Healthcare, Legal, Insurance, Public Sector, and Pharma. Built an async Document Intake and Review Platform using Cloud Storage and Pub/Sub workers for OCR processing and metadata normalization at million-page scale. Implemented source-grounded RAG workflows with Vertex AI (Vector Search, BigQuery) enabling Gemini summarization with lineage and audit-ready traceability. Designed passage-level source linking to improve factual grounding and reduce reviewer cycle time. Hardened async production systems with improved retry/failure recovery and added analytics/monitoring for reliability. Supported GKE canary deployments and model/version rollout checks to ensure zero unplanned downtime.

Senior Member of Technical Staff, Einstein AI Platform at Salesforce

Built and maintained backend APIs and an AI request governance layer for the Einstein conversational AI platform, including routing, metering, and rollout infrastructure for enterprise AI workloads at high request volumes. Developed request governance services for API entry, prompt routing, usage capture, and event processing. Created enterprise monitoring dashboards for real-time visibility into usage, billing, and model performance. Designed SQL/NoSQL backed usage stores to support billing alignment, quota enforcement, rate limiting, and idempotent retry handling. Implemented observability tooling with structured logging, distributed tracing, and Prometheus/Grafana using OpenTelemetry. Managed code generation and model releases with versioning, canary rollouts, and feature launches while preventing duplicate billing events.

Software Engineer, Analytics Intelligence at Tableau

Built the backend product stack behind Tableau Explain Data, including AI scoring pipelines, dashboard query integrations, and a serving layer delivering ranked data explanations for outliers, value spikes, and unusual patterns. Developed backend services for generating ranked candidates with interactive p95 latency under 2 seconds. Designed explanation scoring and model pipeline with SQL processing, Bayesian inference, ML scoring, and probabilistic ranking. Bridged data science prototypes to production analytics features by connecting model outputs with Tableau dashboards and query engine context. Optimized UI-backend paths with caching and WebSockets for real-time explanation rendering performance at dashboard scale; tuned query plans and sampling strategies for interactive latency targets.

Software Engineer Internship at Under Armour