UAE

Pushker Sahai

Sovereign AI & Data Platform Leader

Google (San Francisco) · Microsoft (Seattle) · Cloudera (Toronto)

I help enterprises design and deliver governed AI and modern data platforms—secure, audit-ready, and scalable across cloud and hybrid environments.

18+ yearsEx GoogleEx MicrosoftEx ClouderaCTO: Multi-agent AI (100K+ concurrency)Regulated enterprises

What I do

Sovereign AI & Governance

Data residency, access control, model lifecycle governance, auditability, and Responsible AI aligned to regulated environments.

Modern Data Platforms

Warehouse/lake/lakehouse architectures with lineage, quality, and operating models that unlock trusted analytics and AI.

AI Engineering at Scale

Production RAG + multi-agent orchestration, LLM evaluation/observability, LLMOps, and performance/cost optimization.

Featured work

CTO: Multi-Agent Orchestration Platform

Led architecture and execution for a multi-agent orchestration layer built for reliability and horizontal scale (designed for 100K+ concurrent users).

Multi-Agent OrchestrationMicroservicesLLMOpsReliability/Scale

Enterprise AI & Data Reference Architectures (Google)

Drove executive workshops and reference architectures for governed AI and modern analytics—accelerating POV → production with security and standards.

BigQueryVertex AILookerGovernance/Security

Enterprise Modernization & Delivery (Microsoft)

Directed modernization roadmaps and delivery execution across cloud/hybrid estates—aligning architecture, operating model, and compliance for scale.

Azure MLDatabricksCloud ModernizationSecurity/Compliance

Hybrid Architecture & Governance (Cloudera)

Designed regulated-ready warehouse/lake/lakehouse options and guided POC/POV lifecycles to production adoption with governance and lineage.

Hybrid CloudGovernance/LineageLakehouseEnterprise Architecture

How I help

Sovereign AI Readiness (2–3 weeks)

Assess target use cases, data residency constraints, and governance gaps. Deliver a pragmatic control framework and reference architecture.

  • Use-case shortlist
  • Governance controls
  • Reference architecture
  • Implementation plan

Data Platform Modernization Blueprint (3–6 weeks)

Define target architecture and migration waves for a governed warehouse/lake/lakehouse foundation that supports AI at scale.

  • Target architecture
  • Migration waves
  • Operating model
  • Success metrics

AI to Production Accelerator (4–8 weeks)

Stand up production patterns for RAG/agents with evaluation, observability, security guardrails, and deployment workflows.

  • LLMOps patterns
  • Eval + observability
  • Security guardrails
  • Pilot-ready system

Selected engineering artifacts