AI Tooling Suite — MCP Servers & AIRIA
The challenge
Bring internal engineering systems into safe AI context without exposing sensitive enterprise data.
The architecture
Built MCP servers for Azure DevOps, Active Directory, and localization workflows, then orchestrated AIRIA agents through RAG pipelines, guardrails, and intent-based tool routing.
What I did
-
Designed and delivered a suite of MCP servers bringing SDD's internal systems into AI agent context: Azure DevOps Server, Active Directory, and TransLynk localization platform.
-
AIRIA Orchestration: engineered AI agents, RAG pipelines, security guardrails, and intent-routing models within SDD's enterprise environment for policy-compliant agentic automation.
-
Azure DevOps MCP Server: exposes on-prem Azure DevOps Server 2022.2 as a structured context source for AI agents — covering projects, repositories, pipelines, work items, and build history.
-
TransLynk MCP Server: scans application source code to extract all localization strings, auto-generates EN/AR translations, diffs against existing translations, and upserts only net-new entries.
-
Active Directory MCP Server: structured query interface into on-prem AD — enabling identity lookups, group membership queries, and org-chart context resolution within agentic workflows.
-
AIRIA Orchestration Platform: engineered AI agents, RAG pipelines, security guardrails, and intent-routing models within SDD's enterprise AIRIA environment for policy-compliant agentic automation.
-
OnePortal AI Assistant: context-aware chatbot using AIRIA RAG pipelines — ingesting GRC policies, HR policies, IT help content, and the OnePortal service catalogue.
The impact
Cut manual operational lookup time, accelerated engineering workflows with context-aware AI actions, and maintained policy-aligned access boundaries.
Technologies
- .NET
- MCP Server SDK
- Azure DevOps REST API
- LDAP / Active Directory
- AIRIA Platform