RegIQ is a multi-agent regulatory compliance intelligence system built for the Band of Agents hackathon. When a new regulation arrives, five specialized AI agents collaborate in a cascade pipeline: the Monitor assesses urgency, the Legal Parser extracts requirements, the Impact Mapper maps them to company SOPs via RAG (Qdrant), the Gap Analyst identifies compliance gaps, and the Remediation Planner generates prioritized tickets with dependency chains. The final report goes to a human reviewer who approves or rejects it through a focused HITL interface. Built on the Band SDK with model routing across Gemini, Claude, and GPT-4o, RegIQ demonstrates how orchestrated agents can compress weeks of compliance work into minutes while keeping humans in control of the final decision.
Category tags:"RegIQ - Scoring **Application of Technology** • Criteria: Application of Technology • Score: 4 • Rationale: 5-agent cascade pipeline is solid. Band SDK with model routing (Gemini, Claude, GPT-4o) shows multi-model approach. RAG (Qdrant) for SOP mapping is good technical choice **Presentation** • Criteria: Presentation • Score: 4 • Rationale: Clear documentation, GitHub + PDF + demo available. Workflow clearly explained **Business Value** • Criteria: Business Value • Score: 5 • Rationale: Regulatory compliance is massive market. Weeks to minutes claim is valuable. Every company needs this **Originality** • Criteria: Originality • Score: 4 • Rationale: RegTech + multi-agent is a strong combination. HITL interface keeps humans in control - important for compliance Total: 17/20 🚀 Feedback - RAG integration for SOP mapping is practical - Final human review is crucial for compliance use cases"
Sanem Avcil