About the Role

We are looking for a Senior AI Architect to own the end-to-end technical architecture of a production-grade, multi-agent AI platform deployed within US-regulated financial institutions.

This is a hands-on architecture role, not an advisory or strategy position. The architect will be responsible for critical design decisions across AI/ML platforms, cloud architecture, security, compliance, scalability, deployment, and operational reliability.

The role involves working with Azure and GCP, Kubernetes, BYOC deployments, confidential computing, agentic AI, RAG, LLMOps/MLOps, identity and access management, encryption, and immutable audit architecture.

Key Responsibilities

1. Platform Architecture

- Own the overall architecture of a multi-agent AI platform across Azure and GCP.
- Define component boundaries, data flows, integration patterns, and deployment topologies.
- Architect the Model Abstraction Layer for multi-model and multi-cloud flexibility.
- Design Blue-Green deployment strategies, traffic shifting, rollback, and regression gates.
- Design scalable AKS/GKE Kubernetes architectures including namespace isolation, node pools, workload identity, and resource management.

1. Security Architecture

- Architect BYOC / tenant-isolated deployments ensuring customer data remains within the client's cloud environment.
- Design offline cryptographic licensing using RSA/JWT-based mechanisms.
- Define encryption architecture using Azure Key Vault / GCP Cloud KMS, CMEK and TLS 1.3.
- Design confidential-computing architecture for secure data processing.
- Architect private networking using Azure Private Link/Endpoints and GCP VPC Service Controls/Private Service Connect.
- Design Entra ID / Google Cloud Identity, SSO, Workload Identity Federation and RBAC.
- Define immutable, tamper-evident audit logging covering agent actions, HITL decisions, evidence retrieval, and output versions.

1. AI System Architecture

- Architect multi-agent / agentic AI pipelines, workflow orchestration, agent handoffs, HITL gates, and failure handling.
- Design RAG architectures, including embeddings, vector databases, chunking, indexing, retrieval optimization, and context management.
- Design deterministic tool architectures separating AI reasoning from data retrieval.
- Ensure tool access, evidence validation, errors,



and data operations are fully auditable.
- Architect stateless and reproducible agent execution.

1. Cloud & Infrastructure Architecture

- Design multi-cloud architectures across Microsoft Azure and Google Cloud Platform (GCP).
- Architect persistence using Azure SQL / PostgreSQL, Azure Blob Storage / GCS.
- Design cloud-native observability using Azure Monitor, Application Insights, Cloud Monitoring and Cloud Logging.
- Define secure container supply chains using Sigstore, Binary Authorization, Azure Container Registry and Google Artifact Registry.
- Establish Kubernetes admission and image-signing controls.

1. Enterprise Integration Architecture

- Design integrations with ERP, IAM, GRC and workflow platforms across on-premise and cloud environments.
- Architect GRC integrations using APIs and file-based interfaces.
- Define data mappings and output formats including JSON, PDF and CSV.
- Design deployment packages using Helm, Terraform and signed container images.
- Establish secure, zero-Anaptyss-access deployment processes.

Required Experience & Skills Architecture & Cloud
- 8+ years of enterprise software architecture experience.
- At least 3 years of AI/ML platform architecture experience.
- Experience designing production systems in regulated environments, preferably financial services.
- Strong cloud-native architecture experience with Azure and/or GCP.
- Strong knowledge of Kubernetes, IAM, encryption, networking and managed cloud services.
- Experience with BYOC / tenant-isolated deployment models.
- Architecture-level experience with confidential computing such as Azure Confidential Computing or GCP Confidential VMs.

AI / ML Architecture
- Strong understanding of LLM inference architecture, model serving, API integration, latency optimization, token economics, and production failure handling.
- Hands-on experience architecting agentic AI / multi-agent systems.
- Experience with RAG architecture,



embeddings and vector databases such as:
- pgvector

- Pinecone

- Weaviate

- Vertex AI Vector Search

- Equivalent technologies
- Strong understanding of LLMOps / MLOps, model versioning, Blue-Green deployments, regression testing, monitoring, and output-quality controls.
- Understanding of LLM security risks including prompt injection, data exfiltration, and context manipulation.

Security & Compliance
- Solid knowledge of RSA, JWT, asymmetric cryptography and key management.
- Experience with cloud security architecture covering:

- CMEK

- TLS

- Secrets management

- Network security

- Confidential computing
- Strong knowledge of SAML/OIDC, SSO, Workload Identity Federation and RBAC.
- Experience designing immutable audit logs and tamper-evident audit trails.
- Familiarity with US banking technology-risk frameworks such as OCC, Federal Reserve SR Letters and FFIEC guidance.

Engineering & Documentation
- Ability to create detailed architecture artefacts including:

- Component architecture diagrams
- Data-flow specifications

- Interface contracts

- Security boundaries

- Deployment topologies
- Experience preparing technical documentation for enterprise architecture, information security, model risk and regulatory review.
- Proficiency in Python and/or TypeScript sufficient to review code and prototype architectural concepts.

Good to Have
- Experience working directly with bank technology risk, information security or enterprise architecture teams.
- Multi-cloud architecture experience across both Azure and GCP.
- Confidential computing implementation experience including attestation, enclave programming or sealed storage.
- Experience designing software licensing and IP protection systems for customer-hosted deployments.
- Experience with SOC 2 Type II audit preparation.

Why Join Us?
- Work on a technically complex enterprise AI platform for regulated financial institutions.
- Solve challenges across agentic AI, LLM architecture, RAG, multi-cloud, confidential computing and cloud security.
- Work closely with senior enterprise stakeholders including bank CDOs, Chief AI Officers and Enterprise Architects.
- Opportunity to influence architecture decisions across a production-grade AI platform.

📌 AI Architect Enterprise AI Platform Architecture & Regulated Systems (Gurugram)
🏢 NLB Services
📍 Gurugram