inTheEU Assist is an on-premise AI agent platform for regulated professional firms. This guide covers every module available in the platform, organized by the drawer groups visible in the application.
Work
Clients
Manage the professional relationships at the center of your workflow. Clients are the top-level entity that cases, documents, and outputs attach to.
Work
Cases
Group documents and AI work into structured cases. Cases are the unit of professional liability and the primary audit boundary.
Work
Documents
Upload, index, and reference source documents. Documents feed knowledge packs and agent context. Ingestion is signed and traceable.
Work
Risk Flags
Structured risk signals raised by agents during case work. Each flag is signed, linked to its source, and requires professional review.
Work
Reviews
The human approval record. Every stage transition that required professional sign-off is stored, signed, and permanently attached to the case.
Work
Generated Outputs
Documents, reports, and structured data produced by agents. Each output carries the signature of the agent, the skill, and the approving professional.
AI and Analytics
Knowledge Packs
Domain knowledge bases assigned to agents and the assistant. Configure what each agent knows. Every chunk is signed and versioned.
AI and Analytics
Adapter Training Sets
Anonymized datasets for domain adaptation. Client identifiers are hashed. Training events are signed. The audit chain covers the model itself.
AI and Analytics
Templates
Reusable prompt and output structures. Templates standardize agent behavior across cases and ensure consistent, auditable outputs.
Agent Runtime
Agents
Build, register, and manage AI agents. Assign skills, knowledge packs, and permissions. Agents can spawn sub-agents within defined limits.
Agent Runtime
Skills
The executable units of agent behavior. Each skill has a prompt template, expected inputs, and outputs. Skills are signed at registration and at execution.