Parth Kasat

About

LLM/agentic workflow specialist with healthcare deployment experience who has taken LLM-based automation from prototype to production using operator-in-the-loop validation, RAG-style retrieval, RBAC, and monitoring for sensitive data compliance. Demonstrated real-time incident resolution (retrieval timeouts due to network/proxy misconfig) and strong GTM support—hands-on developer workshops and sales demos translating technical safeguards and real-time ETL into measurable ROI (70% ops reduction, ~$200K/year savings).

Key Strengths

• Took healthcare LLM workflow automation from prototype to production with secure scaling for sensitive data
• Validated LLM outputs via staging with anonymized data and side-by-side comparison to operator decisions
• Built effective operator-in-the-loop feedback processes (shadowing workflows, mismatch flagging for retraining/prompt updates)
• Real-time debugging of LLM/agent pipelines using logs/metrics/tracing and layer-by-layer isolation
• Stabilized production incidents with pragmatic mitigations (caching, fallback prompts, circuit breakers) while implementing permanent fixes
• Resolved live deployment failure by identifying retrieval timeouts from misconfigured network/proxy settings and restoring service without patient disruption
• Strong technical enablement: runs hands-on workshops with live code and exercises to drive adoption
• Sales partnership: translates technical architecture into business outcomes; supported enterprise healthcare deal with quantified ROI (70% manual ops reduction, ~$200K annual savings)
• Led production deployment of AI workflow automation integrating legacy EHRs and multiple third-party APIs
• Designed staged rollout with parallel environments using Kubernetes and Terraform
• Built real-time ETL pipelines with MLflow while maintaining HIPAA compliance
• Translated regulatory requirements into technical controls (encryption, RBAC, audit logging) in partnership with compliance officers
• Systematic cross-layer troubleshooting (software, hardware, networking) to isolate root cause (misconfigured firewall throttling traffic)
• Delivered customer-specific Python automations/ETL integrations reducing manual operations by 90%
• On-site customer collaboration with aviation operators; iterated dashboards in real time based on operator feedback
• Strong in cloud and backend architecture for scalable pipelines and automated workflows

Experience

Education

Skills

Languages