Akshay

Partnered with data science, compliance, and product teams to gather requirements for GenAI use cases; defined architecture blueprints for LLM-based RAG pipelines; designed autonomous-agent workflows; built scalable retrieval pipelines with FAISS and Bedrock KBs; implemented cross-cloud experimentation with Vertex AI; developed end-to-end ML lifecycle automation, monitoring dashboards, and governance guardrails for secure, compliant AI interactions.

Led AI transformation for financial document automation and client reporting; defined end-to-end LLM architecture with LangChain/LangGraph; designed hybrid retrieval workflows; established prompting, grounding, and dynamic template chaining; built multi-agent orchestration and modular ingestion pipelines; fine-tuned domain-specific models; created federated KB access and context prioritization; implemented CI/CD, MLflow, and cross-cloud model registry with observability and governance.

Conducted EDA on EHR/EMR data, improved data pipelines to boost downstream model accuracy, developed binary and multiclass classifiers, and built sequential models for clinical time-series; implemented HIPAA-compliant inference APIs and dashboards; performed model validation with ROC-AUC and SHAP; ensured FDA/medical software compliance and conducted knowledge transfer.