CLO Dashboard Copilot — Compliance AI

Context

A CLO (collateralized loan obligation) compliance and portfolio-management SaaS, built on Laravel, React and MySQL, wanted a Dashboard Copilot: a conversational layer that answers questions about the portfolio directly from the data already in the platform. This is an ongoing engagement.

The Hard Part

In a compliance context a confidently wrong answer is worse than no answer at all, so retrieval is restricted to approved database data and every response has to be grounded in what was actually retrieved rather than generated from the model's own recall.

What I Built

• Azure OpenAI Dashboard Copilot over the existing Laravel, React and MySQL platform
• RAG-style retrieval architecture scoped to approved database data only
• Structured context payloads assembled per question before the model sees them
• Grounded answers constrained to retrieved data, so the copilot does not hallucinate
• AI guardrails bounding what the copilot will attempt to answer
• Token and cost control on every request

Tech Stack