Projects Showcase

Managed the deployment of a smart-home/office IoT platform on Azure. Leveraged GitHub Actions to automate CI/CD, configured Azure IoT and Azure Front Door, integrated Azure ADX for analytics, optimized app services, and implemented centralized logging and monitoring to ensure performance, reliability, and scalability.

Led the deployment of the Inzer Gym App infrastructure on AWS. Utilized Terraform for resource provisioning, streamlined deployments via GitHub Actions, implemented disaster recovery plans, optimized costs, and ensured high availability. Collaborated with cross-functional teams to fine-tune performance, monitor system health, and proactively address issues.

Played a pivotal part in deploying and optimizing the secure tax office infrastructure. Leveraging AWS services and Terraform, I orchestrated a robust, highly available environment. I streamlined deployments with GitHub Actions, implemented disaster recovery and backup plans, optimized costs, and improved security. By automating routine tasks and implementing vigilant monitoring, the result was a resilient and efficient tax office system.

Architected and maintained high-availability AWS infrastructure for a popular restaurant chain, leveraging EKS for containerised microservices, ensuring 99.9% uptime for online ordering, and catering management systems. Implemented scalable AWS services including Elastic Beanstalk, RDS, ElastiCache, and CloudFront to optimise performance during peak hours, reducing page load times by 40% and supporting 200% growth in online transactions.

Modernised a supply chain and e-commerce fulfilment platform on Azure by implementing an event-driven architecture. Deployed AKS for automated customer-data processing within a network-isolated environment and built an integration module to automate customer onboarding. Integrated a customer support chatbot with core infrastructure using Azure AI Foundry to handle shipment and order queries.

Built an event-driven serverless WhatsApp customer support agent using Azure AI Foundry, AI Search, and Cognitive Services with Meta Business API integration. Implemented multilingual support with intelligent routing, real-time analytics dashboards, and automated escalation workflows. Deployed on Azure Functions with ServiceNow ticketing integration, reducing response times by 60% while maintaining high satisfaction scores.

Orchestrated the deployment of the Cosmic AI infrastructure on Azure using Infrastructure as Code. Streamlined frontend and backend deployments through Azure DevOps, integrated Azure OpenAI services to build an intelligent chatbot, implemented disaster recovery and high-availability plans, and collaborated with cross-functional teams to optimize continuous deployment and scalability.

Developed a real-time voice banking agent using Azure GPT Real-Time API with RAG-based query resolution for customer calls. Built intelligent call routing that resolves queries via AI Search and Azure OpenAI, with automatic escalation to relevant customer support when needed. Deployed on Azure App Services with infrastructure provisioned via Azure Bicep, integrated Azure AI Foundry for enhanced capabilities, and implemented Azure DevOps monitoring that reduced call handling time by 45%.

Developed a comprehensive enterprise AI platform combining multiple specialized agents: voice agent for real-time conversations, search agent for knowledge retrieval, GPT agent for general queries, and RAG chat agent for document-based responses. Built centralized admin management system for user roles, agent configuration, and analytics. Deployed on Azure using AI Foundry, OpenAI services, and DevOps pipelines with scalable architecture supporting multiple enterprise clients.

Established a cloud-native MLOps framework on Google Cloud's Vertex AI for responsible deployment of fine-tuned generative models. Built orchestration with Vertex AI Pipelines and Cloud Build, maintained artifacts in Cloud Storage, and exported metrics to BigQuery. Implemented custom monitoring tracking content-aware metrics (latency P95, refusal rates, citation faithfulness) with automated evaluation via Cloud Scheduler.