Associate Director - AI Platform Architect

Sirius AI; Gurugram District; On-site; INR 4,000,000 - 8,000,000; Full time

Job summary

Sirius AI is recruiting a seasoned Enterprise Architect to lead end-to-end data and cloud-native platform initiatives. You will shape pre-sales to production, design scalable architectures, and build enterprise-grade Agentic AI systems using LLMs, RAG, and multi-agent frameworks for multi-tenant environments.

Responsibilities include defining platform patterns, governance, RBAC, and monetization-ready architectures while driving modern, AI-native transformations across financial services clients.

Qualifications

• 12+ years designing enterprise data and cloud-native platforms.
• 5 years as Principal / Lead Architect on Azure or AWS or GCP.
• Demonstrated experience building and productionizing GenAI or Agentic AI systems.

Responsibilities

• Lead enterprise data platform builds from pre-sales shaping to production deployment.
• Define target architectures across channels, integration, cloud, data, microservices, and security.
• Architect and build enterprise-grade Agentic AI systems leveraging LLMs, RAG, multi-agent frameworks and tool-use architectures.
• Design and implement Agentic Operating Systems enabling multi-agent collaboration and memory management.
• Lead architecture and development of a scalable, multi-tenant Agentic AI product platform.
• Define monetization-ready architecture and platform roadmap for repeatable AI products.

Skills

Enterprise Architecture

Azure/AWS/GCP

GenAI

Multi-tenant SaaS

LLM architectures

AI governance

CI/CD

Consulting experience

Platform engineering

Tools

Azure DevOps

GitHub Actions

Job description

Key Responsibilities

• 1. Enterprise Architecture & Data Platforms Engage with clients to understand their architecture, constraints and business objectives Lead enterprise data platform builds from pre-sales shaping to production deployment Define target architectures across channels, integration, cloud, data, microservices, and security Modernize legacy ecosystems into scalable, AI-native platforms
• 2. Enterprise-Grade Agentic AI Platforms Architect and build enterprise-grade Agentic AI systems leveraging LLMs, RAG, multi-agent frameworks and tool-use architectures Design and implement Agentic Operating Systems (Agent Orchestration Layers) enabling: Multi-agent collaboration Tool and API execution layers Short- and long-term memory management Guardrails and policy enforcement Human-in-the-loop workflows Define reusable agent design patterns and enterprise agent SDK standards Evaluate and productionize frameworks such as LangGraph, Semantic Kernel, AutoGen, CrewAI, etc.
• 3. Multi-Tenant Agentic AI Product Platform Lead the architecture and development of a scalable, multi-tenant Agentic AI product platform hosted on Sirius AI cloud infrastructure Design the platform to support: Secure tenant isolation (data, memory, prompts, agents) Configurable client-level agent customization Role-based access controls (RBAC) Cost attribution & token usage metering per tenant Model routing & provider abstraction layers Define and implement: Tenant onboarding workflows Secure data ingestion & RAG pipelines per client API gateway & SDK layers for enterprise integration White-label and configurable deployment models Establish reusable core agent services that can be extended across multiple Financial Services clients Build a roadmap for evolving the platform into a repeatable AI product offering, not just bespoke consulting delivery Ensure platform scalability, resilience, and high availability across geographies Define monetization-ready architecture (subscription, usage-based, hybrid models)
• 4. AI Governance, Observability & Reliability Design and implement enterprise AI observability frameworks using: Langfuse Opik Weights & Biases Azure AI Studio monitoring Custom telemetry pipelines Establish standards for: Prompt observability & version control Token and cost tracking Model & agent performance monitoring Agent traceability and execution logs Hallucination detection and evaluation pipelines Implement offline online evaluation loops for LLM and agent systems Embed audit logging, compliance, explainability and AI risk controls suitable for Financial Services enterprises
• 5. Cloud, DevOps & Platform Engineering Lead Azure/AWS architecture including Compute, Storage, Networking, AKS/Kubernetes, DevOps, Security and Monitoring Implement MLOps / LLMOps pipelines for model lifecycle, prompt lifecycle and agent deployment Drive performance tuning, resilience engineering, and cost optimization Manage security findings, vulnerabilities and control gaps across applications and infrastructure
• 6. Strategic Advisory & Leadership Act as trusted advisor to CTOs, CDOs and AI leaders Drive GenAI and Agentic AI transformation roadmaps Lead innovation programs, accelerators and reusable IP development Build and mentor high-performance architecture and AI engineering teams Shape Sirius AIs internal AI platform strategy and long-term product vision

Job Requirements

• 12 years designing enterprise data and cloud-native platforms
• 5 years as Principal / Lead Architect on Azure or AWS or GCP
• Demonstrated experience building and productionizing GenAI or Agentic AI systems
• Experience architecting multi-tenant SaaS or platform products
• Hands-on exposure to: LLM architectures (OpenAI, Azure OpenAI, Anthropic, OSS models) RAG systems and vector databases Multi-agent orchestration frameworks Agent memory systems and tool-use design
• Experience implementing AI observability and evaluation frameworks (Langfuse, Opik, etc.)
• Strong understanding of: AI governance & risk controls Prompt lifecycle management Cost optimization for LLM workloads Subscription, self-hosted and hybrid deployment strategies
• Expertise in CI/CD (Azure DevOps, GitHub Actions, etc.)
• Experience in consulting services preferred
• Experience building reusable AI accelerators and platform IP strongly preferred

Benefits

Work with a very innovative and collaborative firm focused on harnessing the power of AI and clients data for cutting edge solutions. Founders are