AI Technology Enabler FDE Cloud Security Operations

Abu Dhabi Dubai UAE India

Your trusted builder for Cloud and AI.

Kunal is a Technology Enabler with hands-on experience building, deploying and operating at scale across Cloud and AI. He embeds with the business unit, finds the use case actually worth building, and carries it from pilot to production — most recently seven GenAI services taken live across Abu Dhabi Government, all demoed at GITEX 2025.

Years in technology
0+
Cloud & AI projects
0+
Cloud & AI certifications
0+
AI services in production
0+
GenAI solutions delivered
0+

Kunal builds things that have to keep working.

Kunal is a multi-cloud and AI certified solution architect with more than a decade of building behind him, currently an AI Technology Enabler and forward deployed engineer with the Federal Government in Abu Dhabi, and a founding member of its AI Factory team. Most of what he builds now is agentic — agents with real tool access, and the agent harness around them that decides whether they are dependable or merely impressive.

The job is forward deployed in the real sense: he sits with the department, works out which foundation-model use case survives contact with reality, and then builds, secures, ships and operates it. Seven flagship services went from pilot to production that way, spanning conversational assistants, contextual RAG platforms and workflow automation — and all seven were demoed at GITEX 2025.

Before AI took over the work, he spent a decade on cloud that has consequences: finance and licensed financial institutions, government, oil and gas, healthcare and AI imaging, public safety. Migrations that cannot drop a transaction, DR that gets tested for real, bills that quietly doubled. That grounding is why the AI systems he builds have audit logging, prompt filtering and data residency in them from the first commit rather than after the first incident.

Based
Abu Dhabi, UAE
Role
AI Technology Enabler
Focus
Agentic AI & platform

What Kunal actually gets called in for

01 / 05

Scroll or use ← →

AI ENGINEERING

Getting a foundation model to do something a department will rely on.

  • RAG done properly — contextual, graph and agentic, not just embeddings in a vector store
  • Agentic workflows with tool use, and the orchestration to keep them predictable
  • LLM application platforms: LangChain, LangGraph, LangSmith, LiteLLM, Phoenix
  • Real-time voice pipelines — Whisper ASR, Azure Speech streaming, Hamsa

AI PLATFORM

The unglamorous layer that decides whether inference is affordable.

  • GPU inference architectures on Kubernetes with vLLM-powered serving
  • Model routing gateways and high-throughput deployment strategies
  • Vector and graph stores — Milvus, Neo4j, Zep, LightRAG, GraphRAG
  • Secure service exposure via FastAPI, API Gateway and private networking

AI GOVERNANCE

What makes AI deployable in a regulated environment at all.

  • Audit logging and prompt filtering as platform features, not bolt-ons
  • Data residency and sovereignty enforcement for public-sector workloads
  • Access management and tenancy boundaries across government entities
  • Responsible-AI controls that survive an actual compliance review

CLOUD ARCHITECTURE

Designing it before it becomes expensive to change.

  • Landing zones, Control Tower, multi-account strategy and org guardrails
  • Migration design — assessment through cutover runbook
  • DC-DR architecture and BCP that has been failed over at least once
  • Well-Architected reviews that produce a decision log, not a scorecard

PLATFORM & DELIVERY

So the next team does not have to repeat any of this.

  • Terraform and CloudFormation, with a review culture around them
  • GitOps pipelines — Jenkins, GitHub Actions, Azure DevOps, ArgoCD
  • Observability that answers questions: Dynatrace, Grafana, Datadog, ELK
  • FinOps — attribution before cutting, then rightsizing and commitments

Builders project list

MVP to production, every time — not stuck at demo.

  • Government · GenAI

    Government-scale AI assistant

    Conversational assistant built on Azure with RAG and agentic pipelines, taken from pilot to production and demoed at GITEX 2025 as part of a sovereign AI enablement programme.

    • Azure
    • RAG
    • Agentic AI
    • FastAPI
  • Government · Platform

    AI Factory

    Founding core member. Identifying, validating and operationalising strategic foundation-model use cases across a government ecosystem — the pipeline that turns a department idea into a running service.

    • Use-case discovery
    • MLOps
    • Governance
  • Kubernetes · vLLM

    GPU inference platform

    Scalable GPU-backed inference on Kubernetes with vLLM serving, model routing gateways and high-throughput deployment strategy, sized so that cost per token stays defensible.

    • Kubernetes
    • vLLM
    • NVIDIA GPU
    • HPC
  • Confidential client · AI Platform

    GenAI platform & GPU fleet operations

    End-to-end GenAI platform: model deployment and inferencing, AIOps and model observability, and orchestration of H100 and H200 fleets sized so cost per token stays defensible under production load.

    • H100 / H200
    • Inferencing
    • AIOps
    • Model observability
  • Real-time · Speech

    Voice AI assistant pipelines

    Streaming voice assistants integrating Whisper ASR, Hamsa and Azure Speech with conversational orchestration, built for latency low enough to feel like a conversation.

    • Whisper
    • Azure Speech
    • Streaming
  • Healthcare · AI Imaging

    Healthcare imaging platform

    Teleradiology pipeline handling DICOM ingest through to reporting, built to keep patient data resident and auditable end to end.

    • AWS
    • DICOM
    • VPC design
    • Compliance
  • Finance · LFI · Oil & Gas

    Regulated cloud migrations

    Datacentre-to-cloud migrations for licensed financial institutions and industrial operators, where the recovery objective is contractual and the DR posture is exercised on a schedule rather than assumed.

    • Migration
    • DC-DR
    • Multi-account
    • Audit

Words from clients and colleagues

Unedited recommendations from the people who worked alongside the delivery.

6 recommendations

He brings a rare combination of deep technical expertise and strong architectural thinking. What sets him apart is his focus on building technology the right way — with attention to governance, security, reliability, and long-term sustainability, not just quick wins.
Piyush Mathiya Founder, Innovatrae Solutions Was Kunal’s client ·
Kunal is extremely knowledgeable and consistently provided support whenever we faced challenges with AKS. His guidance made a significant difference, especially during complex or time-sensitive situations. What stood out the most was his calm approach and clear communication.
Sarith Sasidharan Microsoft Cloud Security, Identity & Governance Was Kunal’s client ·
Kunal was instrumental in delivering scalable, reliable AWS solutions tailored to our customers. His technical depth combined with a collaborative spirit made complex deployments feel seamless. Kunal’s calm under pressure and proactive communication earned our full trust.
Nishanth Shashidhar Founder, neuwork Same team ·
Kunal’s proficiency in AWS services and Docker/Kubernetes are un-parallel to anyone I have ever worked with before. His technical expertise and problem-solving skills were invaluable, enabling us to overcome complex challenges and deliver outstanding results.
Mayuraksha Sikdar AVP Data Engineer, JPMorgan Chase Same team ·
He played a crucial role in designing and implementing our AWS infrastructure, architecting scalable and efficient solutions that significantly improved our deployment processes and enhanced our overall system reliability. He was an invaluable asset to our organization.
Deepa Narra VP, Lead Data Scientist, JP Morgan Chase Same team ·
I was particularly impressed by his technical skills and ability to handle even the toughest client. He goes beyond in solving customer problems and ensures that we meet customer expectations.
Ankita Singh Engineering Manager, Software Definition Managed Kunal’s team ·

Tell Kunal what you are trying to ship.

An AI use case that has stalled at pilot, a RAG system that is confidently wrong, inference costs that do not add up, or a cloud estate that needs to stop paging people. Short note is fine.