Hemachandar

Principal ML Platform & AI Infrastructure Engineer with 15+ years of experience leading the design, delivery and operation of large-scale production platforms across AI/ML, cloud infrastructure and mission-critical financial systems. Currently driving AI platform and infrastructure strategy at UBS, combining hands-on engineering depth with technical leadership across Kubernetes, GPU infrastructure, LLM systems, distributed workflows, cloud platforms and observability. Experienced in defining platform architecture, establishing engineering standards, mentoring engineers, coordinating cross-functional teams, and turning ambiguous AI requirements into secure, scalable and reliable production capabilities.

Proven ability to lead initiatives from strategy and architecture through implementation, production adoption, and continuous improvement. Particularly focused on building reusable platform primitives that allow AI engineers and product teams to experiment rapidly while maintaining strong standards for reliability, security, observability, cost and operational excellence.

Principal Platform & AI Engineer (Trading & AI Infrastructure) at UBS AG (2018-08 – Present)

• Provided technical leadership for AI/ML platform strategy, defining architectural patterns and reusable infrastructure capabilities for production of AI workloads across engineering teams.
• Led cross-functional delivery across AI engineers, researchers, platform engineers and product teams, translating evolving AI requirements into scalable, production-ready systems.
• Established engineering standards and reference architectures for Kubernetes, GPU workloads, LLM inference, observability, security and deployment automation.
• Mentored engineers across platform and AI engineering disciplines, raising standards for system design, reliability, automation and production operations.
• Drove platform standardization through reusable Terraform modules, Helm charts, Kubernetes patterns and GitOps workflows, reducing duplicated engineering effort and accelerating adoption of AI capabilities.
• Owned technical direction for complex AI infrastructure initiatives from architecture and proof of concept through production rollout, operationalization and continuous improvement.
• Architected and operated GPU-accelerated Kubernetes infrastructure in Azure for UBS AI platforms, supporting production LLM training and inference workloads with multi-region resilience and enterprise security controls.
• Designed production LLM workflow infrastructure integrating Azure OpenAI, including secure model access, prompt orchestration, structured outputs, hallucination mitigation and failure handling.
• Built reliable long-running AI workflows using ordered task queues, background orchestration, intermediate state management, token lifecycle handling and secure request propagation under concurrent workloads.
• Engineered real-time LLM response streaming using SignalR/SSE with structured delta handling, state management and end-of-stream detection for reliable production delivery.
• Designed end-to-end data and schema contracts across frontend, backend and AI services, enabling consistent handling of complex financial data and structured LLM inputs/outputs.
• Established production observability using Prometheus, Grafana, OpenTelemetry and Azure Monitor, defining SLIs/SLOs around latency, throughput and reliability and improving incident diagnosis and MTTR.
• Developed domain-aware AI failure handling for infeasible workflows, unavailable data, constraint violations and malformed outputs, enabling systematic detection and diagnosis.
• Acted as a technical advisor to engineering teams on cloud architecture, Kubernetes, distributed systems, AI infrastructure and production reliability.
• Built and maintained Python-based automation and orchestration components supporting AI platform operations and workflow execution.
• Designed and operated Kubernetes clusters and CI/CD pipelines in GitLab to power mission-critical, low-latency applications ensuring compliance, resilience, and performance and GPU-heavy AI/ML pipelines (UBS STAAT AI).
• Built Terraform and Helm charts, Ansible-style modules with YAML-based automation to deliver secure, reproducible environments across multiple teams.
• Supported highly available Postgres/SQL DB clusters improving availability for analytics and risk management.
• Architected multi-region failover and resilience strategies to guarantee uninterrupted operations.
• Partnered with cross-functional stakeholders (developers, compliance, and operations) in Agile delivery models, embedding DevOps best practice across the organization.
• Established an observability-first culture, embedding Prometheus, Grafana, and OpenTelemetry into clusters with defined SLIs/SLOs for latency, throughput, and error budgets, reducing MTTR significantly in production.
• Mentored and led a team of 10+ engineers in AI, DevOps and Platform Engineering.

DevOps & Platform Engineering at Bank of America (2015-11 – 2018-08)

• Led the adoption of DevSecOps practices, incorporating static code analysis tools like SonarQube into CI pipelines, improving code quality and security standards across the organization.
• Developed Python automation scripts for SFTP file transfers and data processing, improving operational efficiency.
• Maintained 100+ Autosys Jobs in UNIX, ensuring seamless job scheduling, improving batch processing by 25%
• Directed the research and proof of concepts for onboarding new DevOps tools
• Implemented Kafka for real-time data ingestion, enhancing the analytics capabilities, and reducing data latency.

QA Engineer – Capital Markets at Fidelity Investments (2011-02 – 2015-11)

• Worked on industry critical Security Master Reference data and Pricing platform migration from Mainframe
• Ensured reliability of securities pricing and reference data platforms critical to capital markets trading.
• Built automated test frameworks in UFT, Test Complete boosting test coverage by 50%
• Partnered with developers to resolve latency and data integrity issues, reducing operational risk.
• Collaborated with production teams to improve resilience and failover of trading data systems.

Bachelor of Technology in Computer Science & Engineering – India (2010-07)