Full-Stack / Azure Engineer — Review AI Agent Platform + Build Dashboards & UI

Dear Freelancers, We're a European real estate investment and development firm building an internal AI agent platform on a Microsoft 365 / Azure stack. The core retrieval and agent logic is already in place and we're now looking for an experienced engineer to (1) conduct an independent architecture and security review of the existing setup, and (2) design and build a set of dashboards plus a polished user interface on top of it. The existing setup (high level) Multi-agent fleet architecture (business-level and project-level agents), where "project" is a runtime parameter rather than hard-coded per agent Azure OpenAI for generation; SharePoint / Microsoft Graph as the retrieval and permissions source of truth Azure Functions back-end with an on-behalf-of (OBO) identity model and a web-app front-end Entra ID for authentication, security groups, and access control Document ingestion with a managed metadata layer (tagging, sensitivity classification, status tracking) What we need Independent review — assess the current architecture, retrieval approach, identity/permission model, and security posture. Deliver a written findings report with prioritized recommendations (correctness, security, performance, maintainability). Dashboards — build operational dashboards covering things like agent/fleet status, evaluation results, ingestion and onboarding progress across multiple data rooms, document counts, and tagging/metadata coverage. User interface — design and implement a clean, responsive front-end for end users to query the agents and view results with source citations, respecting the existing OBO permission model (users only see what they're entitled to see). Required skills Strong React / TypeScript, with real data-visualization / dashboard experience Azure: Functions, App Service, Azure OpenAI, Key Vault / managed identity Microsoft Graph API and SharePoint integration Entra ID authentication (MSAL, OBO / delegated flows) Experience with LLM / RAG applications Security-conscious mindset around access control and data handling Nice to have Experience with regulated or confidentiality-sensitive environments German language (working proficiency) — not required but a plus Familiarity with evaluation/observability for LLM systems Deliverables Architecture & security review document with prioritized findings Working dashboard(s) deployed to our environment User-facing query UI, documented and handed over with source code