Shivakumar Appadurai

Hands-on engineering leader responsible for architecting, building and operating the Azure-based LLM Platformand Java Spring
Boot microservices powering JPMorgan Chases enterprise AI/ML initiatives across global Lines of Business.
• Platform Architecture: Architected and delivered a scalable, enterprise-grade LLM Platform on Azure, enabling 5, 000+
data scientists across all JPMorgan Chase Lines of Business to access, fine-tune and deploy large language models
securely.
• LLM Infrastructure: Designed and built multi-tenant LLM infrastructure leveraging Azure OpenAI Service, Azure Cognitive
Services, Logic Apps, Key Vault, and Storage Accounts, achieving 99.9% availability and sub-300ms P95 latency at scale.
• AI Evaluation: Developed Python-based LLM evaluation scripts and model observability pipelines to track response
quality, token usage, latency and model drift, enabling data-driven decisions on model upgrades and fine-tuning.
• Microservices Engineering: Hands-on engineered Java Spring Boot microservices to expose LLM capabilities as secure,
versioned REST APIs, enabling seamless integration by 20+ downstream product teams across Investment Banking, Asset
Management and Retail Banking.
• Security & Compliance: Implemented a secure, policy-compliant LLM gateway with token-level access controls and audit
logging, meeting financial-grade security and regulatory requirements.
• Team Leadership & DevOps: Built and led a high-performing team of 15+ engineers; drove Agile practices, OKR-based
planning, CI/CD automation with Azure DevOps and Terraform IaC, reducing deployment cycle time by 60%.

• Developed and maintained enterprise Java applications; coordinated with onsite teams for requirement clarification and
optimised database-heavy modules, improving query response times by 25%.

• Delivered full-SDLC Java application development including design, testing, code reviews and production deployment;
mentored junior developers on OOP best practices.

• Led a team of 20 engineers delivering Java / Spring Boot microservices for BFSI and retail clients and drove POC-to
production delivery, CI/CD pipelines and code review practices, reducing incidents by 35% and improving release velocity
by 40%.
• Architected distributed microservice solutions using Spring Boot and REST APIs, transitioning monolithic legacy systems
to modular, independently deployable services.

• Participated in all phases of SDLC comprising of analysis, development, testing and code optimization for client-facing
web applications.