Hi, I'm Dhia Gharsallaoui

Technical lead and architect for a multi-tenant GPU cloud platform that extends managed Kubernetes into sovereign and edge datacenters at a scale of thousands of GPUs — owning the stack end-to-end, from cross-site GPU networking and node provisioning to the agentic infrastructure and AI-driven operations layers on top.

• Architected an end-to-end multi-tenant platform running customer AI and GPU workloads on Kubernetes worker nodes hosted in sovereign and edge datacenters, joined to a centrally managed control plane over private connectivity at a scale of thousands of GPUs
• Led the adoption and integration of upstream open-source Kubernetes projects (Unbounded, AKS Flex Node) as the platform's node-layer foundation, letting customer bare-metal and virtual machines join as fully managed worker nodes with automatic GPU enablement and zero-downtime upgrades; built the validation environment that proved cross-site node join and GPU workload scheduling ahead of rollout
• Designed the private connectivity and network architecture linking the control plane and customer environments to sovereign and on-premises datacenters, with strict per-tenant isolation and private-only access to managed services, delivered as modular, staged infrastructure-as-code
• Built the cloud control-plane landing zone on a multi-subscription, GitOps-driven architecture with workload identity and least-privilege, just-in-time access; stood up an AI gateway providing governed, cost-controlled, and safety-filtered access to large language models as a shared platform capability
• Pioneered an agentic Infrastructure-as-Code model in which AI coding agents generate, validate, and evolve cloud infrastructure from natural-language intent with no hand-written infrastructure code — inverting the traditional IaC expertise bottleneck and accelerating safe infrastructure change
• Created an agentic operations (NOC) assistant for AI-assisted detection and triage, exposing curated, read-only, fully audited telemetry tools to a per-customer AI agent runtime built on Model Context Protocol (MCP) servers, pairing agent autonomy with strict security and access controls
• Set technical direction and engineering standards across the platform portfolio — GitOps, test-driven development, security-by-default, and an AI-agent-native development workflow — driving architecture reviews and mentoring engineers across interdependent initiatives

Building production-grade infrastructure for AI-as-a-Service platform, orchestrating multi-cloud GPU compute for training, inference, and AI workloads at scale.

• Architected and built stable, zero-maintenance Kubernetes infrastructure spanning multi-cloud and bare metal environments (AWS, Azure, Sesterce, Scaleway)
• Set up and configured FluxCD for GitOps-based continuous deployment, enabling declarative infrastructure management and automated reconciliation across clusters
• Created custom Helm charts for application deployments, standardizing packaging and configuration management across multi-cluster environments
• Designed and delivered FlexAI-managed customer-facing product enabling "Bring Your Own Compute" on AWS and Azure, defining architecture, building core services, and streamlining deployment workflows
• Brought up, configured, and performance-tuned GPU-based compute infrastructure (Nvidia A10G, H100, AMD MI300), ensuring high reliability and throughput across distributed AI workloads
• Built gRPC services for cross-cluster workload scheduling, enabling efficient orchestration and coordination of AI jobs across geographically distributed Kubernetes clusters
• Designed and implemented custom Kubernetes operators for workload management, automating complex orchestration patterns for AI training and inference jobs
• Integrated and managed GPU Kubernetes operators including nvidia-device-plugin and ROCm gpu-operator, ensuring proper resource allocation and comprehensive monitoring of GPU utilization
• Investigated and deployed low-latency networking solutions and edge cluster serving (Skupper) for distributed AI workloads
• Contributed high-scalability data layer with cross-cloud/region support using JuiceFS for distributed storage across heterogeneous infrastructure
• Established SRE best practices by building robust observability pipelines and monitoring systems, delivering actionable metrics and distributed tracing to accelerate debugging and improve service reliability
• Collaborated closely with runtime teams to ensure AI workloads run smoothly, addressing infrastructure bottlenecks and optimizing resource utilization
• Actively reviewed design documents across teams to ensure cross-functional and business alignment on architectural decisions

Led backend architecture and infrastructure operations for an on-chain trading platform serving DeFi protocols and traders.

• Designed and implemented complete infrastructure architecture from scratch, including network topology, security policies, and scalability frameworks
• Provisioned multi-cloud infrastructure using Terraform across Digital Ocean and OVHCloud platforms, ensuring cost optimization and geographic redundancy
• Integrated Ansible for automated configuration management and deployed Kubernetes clusters with ArgoCD for GitOps-based continuous deployment
• Architected and implemented high-performance trading backend systems in Golang, handling real-time price feeds and order execution
• Built robust on-chain data ingestion pipelines processing millions of transactions daily with sub-second latency requirements
• Established system architecture patterns for high-availability trading infrastructure with zero-downtime deployments
• Implemented comprehensive monitoring and alerting systems for trading operations and infrastructure health

Designed and implemented scalable backend systems and cloud-native solutions with focus on distributed architectures.

• Architected and developed high-performance microservices in Go using domain-driven design principles
• Built event-driven backend systems using Apache Kafka and NATS, processing millions of messages daily
• Designed and maintained multi-cluster Kubernetes infrastructure supporting 50+ microservices
• Implemented comprehensive API architectures with REST and gRPC services achieving 99.9% uptime
• Developed stream processing pipelines using Apache Spark for real-time data analytics and transformation
• Built monitoring and observability systems using Prometheus, Grafana, and distributed tracing
• Established database optimization strategies with PostgreSQL and Redis, reducing query times by 60%
• Provisioned and managed cloud infrastructure using Terraform on Digital Ocean and OVH
• Implemented GitOps workflows using ArgoCD for automated deployments and configuration management
• Built comprehensive CI/CD pipelines reducing deployment time from hours to minutes
• Implemented SSO integration and security hardening practices for production systems

Specialized in blockchain backend development and cross-chain protocol implementation.

• Architected and implemented multi-chain bridge protocols in Go connecting EVM-compatible chains with XRPL
• Developed smart contracts in Solidity and deployed across multiple EVM blockchains
• Built comprehensive backend systems for smart contract interactions, including automated deployment pipelines
• Implemented cryptographic validation systems for cross-chain asset transfers and multi-signature operations
• Designed blockchain-agnostic APIs enabling seamless protocol deployment across multiple networks
• Built robust transaction processing engines with advanced retry mechanisms and comprehensive error handling
• Developed real-time blockchain monitoring systems with alerting for transaction failures and network issues
• Contributed to RWAFi protocol development, building smart contracts in Solidity for institutional-grade real-world asset infrastructure that bridges DeFi liquidity with traditional finance security
• Worked on blockchain wallet extension development, adding EVM-based blockchain support to enhance multi-chain wallet functionality and cross-chain asset management

Developed scalable data ingestion and processing systems.

• Developed scalable data ingestion pipeline using Python/Flask and PostgreSQL
• Implemented containerized microservices using Docker Compose and Kubernetes
• Created automated testing and deployment pipeline using GitHub Actions
• Designed real-time metrics collection system