← William Delo

William Delo

Software Engineer Multi-tenant backend systems AI agent authorization

Software engineer at Microsoft since 2023, building multi-tenant backend systems and authorization for Security Copilot, a generative AI security platform. Built and owned the routing service that maps thousands of customer workspaces to 30+ isolated deployment cells across 8 geographies, enabling the move to a cell-based architecture. Designed the RBAC model that lets teams across 3 other Microsoft Security products assign roles to the AI agents they build. Created the workspace microservice that became a shared dependency across 20 services, and contributed to a team effort that cut median platform overhead 48%.

Experience

Microsoft

Security Copilot, a generative AI security platform

Software Engineer II

2025 – Present
  • Built and owned the routing service mapping thousands of customer workspaces to 30+ isolated deployment cells across 8 geographies, enabling a production move to a cell-based architecture
  • Designed a new SQL-backed RBAC model for AI agents on the platform, combining built-in roles with permissions mapped from 3 other Microsoft Security products so their teams can assign roles to the agents they build
  • Implemented workspace provisioning for Microsoft 365 E5 enterprise onboarding on the new cell-based backend, deterministically allocating each workspace to a cell within its geography
  • Moved workspace reads from an upstream dependency to an owned Azure Cosmos DB store, with fallback and self-healing for workspaces not yet migrated, keeping the service API unchanged for consuming services
  • Created multi-region synthetic monitors, dependency dashboards, alerts, and on-call runbooks for workspace and routing services, enabling earlier outage detection and faster diagnosis of upstream failures

Software Engineer

2023 – 2025
  • Designed and deployed a multi-tenant workspace microservice (C#, ASP.NET Core, Azure Cosmos DB) that became a shared dependency for routing and AI agent features across 20 services, replacing duplicated workspace lookups
  • Instrumented LLM prompt execution with overlap-aware latency tracing and partnered on removing a redundant hot-path call, supporting a team effort that cut median platform overhead 48% (3.2s to 1.7s) and user-perceived latency 14%
  • Fixed a settings API vulnerability by deriving tenant identity from authenticated claims instead of client-supplied request fields
  • Mentored an intern through scoping, design reviews, and final presentation to an accepted return offer

Software Engineering Intern, Intune

2022
  • Containerized a legacy Intune device-management service and deployed it to Azure Kubernetes Service (AKS), validating it with live internal traffic as a proof of concept for migrating existing services to Kubernetes

Skills

Languages
C#, SQL, Kusto Query Language (KQL)
Authorization
RBAC, role and permission modeling, scoped role assignments, multi-tenant isolation, AI agent authorization
Platform and infrastructure
ASP.NET Core, Azure SQL, Azure Cosmos DB, Kubernetes, Azure Kubernetes Service (AKS), cell-based architecture, synthetic monitoring

Education

BS, Computer Science and Engineering, Ohio State University

2023, Summa Cum Laude