AI Infrastructure Architect

Intro: Senior AI Infrastructure Architect responsible for architecting and building custom AI infrastructure and hardware solutions, optimizing performance, power, cost and scalability of the computational stack, advising on technology and vendor selection, and integrating full-stack solutions for production AI/ML applications. Role is based in Chennai and is a full-time position at Accenture. Core responsibilities: - End-to-end architecture and engineering of optimized Snowflake data and AI infrastructure: warehouses, Snowpark workloads, secure data architecture, AI-ready feature/data pipelines, model integration and AI application enablement. - Design and tune scalable Snowflake warehouses, tasks, streams, Snowpark services, Cortex/AI capabilities, Streamlit apps and cloud integrations, including compute sizing, query optimization, governance, access controls and high-throughput data access. - Serve as primary Snowflake AI infrastructure expert; evaluate architecture alternatives and trade-offs across compute, networking, storage, orchestration, model serving, observability, security, compliance and cost. - Lead architecture assessments, identify gaps/risks/bottlenecks, recommend remediation, and document architecture decisions, assumptions and dependencies. - Define capacity planning, scaling strategies, cost forecasts, FinOps inputs and performance improvement opportunities. - Design deployment, automation and CI/CD strategies for reliable, repeatable releases of AI systems, models, data pipelines and platform components into production. - Establish monitoring and observability practices (SLAs/SLOs, alerting, performance/cost tracking) across InfraOps and MLOps. - Integrate AI/ML systems into enterprise environments ensuring interoperability, security, compliance and regulatory alignment. - Mentor engineers, set technical direction, review designs/code and promote engineering best practices. Required qualifications and experience: - Bachelor's in Computer Science, Computer Engineering, Information Technology or related engineering field. - Experience listed as 10-12 years in the job header and the posting also states "Minimum 12 year(s) of experience is required." Additionally, multiple required items cite minimum 4 years in coding/operating AI/ML infrastructure and platform roles. - Strong understanding of AI/ML concepts and infrastructure required to deploy, run and optimize production AI workloads. - Minimum 4 years proficiency in programming/scripting (Python, Java, C++, Bash, PowerShell or equivalent). - Experience with data pipeline/workflow tools such as Apache Airflow, Kubeflow or managed orchestration services. - Proven experience in AI/ML infrastructure engineering on hyperscalers or enterprise platforms and leading AI projects or engineering workstreams. - Strong problem-solving, communication and stakeholder alignment skills. Preferred / Good to have: - Snowflake certifications (SnowPro Advanced Architect, SnowPro Advanced Data Engineer) and hands-on experience with Snowflake governance, secure data sharing, Snowpark, Cortex/AI features. - Exposure to vector search, retrieval pipelines, feature engineering, model enablement and AI application architecture. - Industry experience in regulated sectors (BFSI, healthcare, retail, telecom, manufacturing, energy, public sector) where security, compliance and cost-control are critical. Tools and patterns referenced: Snowflake Data Cloud, Snowpark, Cortex/AI, Streams/Tasks, Streamlit, SQL, Python, dbt, Terraform, CI/CD, DataOps, MLOps, Airflow, Kubeflow, observability and FinOps. The posting emphasizes documentation (architecture decision records, reference implementations, runbooks) and production support operating models.