Jasmit Singh

Sr Engineer specializing in Agentic AI (LangGraph) and Zero Trust infrastructure. Strategic implementer of the Azure Well-Architected Framework to design resilient, secure, and cost-optimized workloads in Azure. Bicep-driven automation, leveraging RBAC, Azure Policies, to govern high-scale environments while maintaining operational excellence through Private AI and Stateful Agent orchestration.

Over 8 years of experience in Enterprise IT.

SKILLS:

AI Frameworks: LangGraph (Stateful Agents), LangChain, Ollama, Llama 3.1, OpenAI, Copilot, Anthropic Claude, RAG.
Infrastructure & Security: Bicep, Zero Trust Architecture, Microsoft Entra ID, RBAC, Conditional Access, Azure Policy.
Azure Compute: VM’s, K8’s, and Azure Container Apps Environment(ACA).
Azure Pillars: Reliability, Security, Cost Optimization, Operational Excellence, Performance Efficiency.
Platform & Monitoring: KQL, Azure SDK for Python, Streamlit, Log Analytics.
Scripting: Python, Bash, PowerShell.

Certifications: Azure Administrator Associate and Azure Solutions Architect Expert.

Projects Only:
Designed and deployed core Azure resources via Bicep templates, ensuring every deployment adhered to the Azure Well-Architected Framework's 5 Pillars. Authored comprehensive Wikis for every deployment, providing a single source of truth for architecture diagrams, security boundary mappings, and troubleshooting steps.

Built production-grade Azure AI Foundry ecosystems using Bicep for automated deployment of models, private endpoints, and Zero Trust networking.

Engineered a stateful "Advisor" using LangChain to orchestrate Azure SDKs and KQL for deep-log analysis (e.g., DeleteBlob/403 errors) via a local Ollama LLM. Integrated interrupt_before human-in-the-loop gates and MemorySaver check pointers to ensure no high-stakes Azure operations execute without manual verification.

Architected and deployed a secure, high-availability Azure Container Apps (ACA) environment within a hub-and-spoke VNet architecture. Implemented (IaC) using Azure Deployment Stacks and Bicep to provision internal-only ingress, private load balancing, and Virtual Network Gateways for secure remote administrative access. Optimized cloud resource lifecycle by managing Container App Revisions, scaling replicas for cost-efficiency, and troubleshooting VMSS-based gateway deployments to ensure 99.9% environment uptime. Running high-performance Ollama inference service on Azure Container Apps to deploy and serve Llama 3.1 (8B) via a private REST API for secure, internal-only LLM consumption with LangGraph Human in the loop.