DPO Radio

A compliance officer, legal counsel, DPO, risk analyst, and auditor may all be working from the same regulations, but they approach the work differently. They ask different questions, rely on different evidence, and prioritize different tools. A single AI assistant tends to answer every question in the same style, regardless of who is asking.
Forseti solves this by providing six specialist AI agent roles instead of one general-purpose assistant. Each persona is a runtime configuration, not a separately trained model. Every persona combines its own system prompt, tool affinity, citation discipline, and refusal logic to match a particular professional role.
All personas work from the same conversation context and the same installed regulatory packs. What changes is how Forseti reasons about the work and which tools it prioritizes. For example, the DPO Advisor naturally favors privacy workflows and transfer assessments, while the Auditor's Eye prioritizes evidence sufficiency, audit lineage, and walkthroughs. The underlying knowledge remains the same.
Personas influence behavior, reasoning style, and tool affinity. They do not determine permissions or access to installed regulatory content. Access is controlled by user permissions, while installed regulatory packs determine the regulatory knowledge available to every persona.
Regardless of persona, Forseti will not invent citations, reach legal conclusions without grounding, accept off-topic redirection into general chat, or present itself as a human. These constraints are built into every persona and remain consistent regardless of which role is selected.
The entry-point persona: fluent across every installed framework
Knows every framework installed in the organisation. Can answer cross-module questions, hand off to a specialist when the question warrants it, and orchestrate multi-step reasoning that spans privacy, risk, and legal work.

Framework-fluent persona for the compliance officer's day-to-day
Owns the framework cross-walk, gap analyses, and control mappings. Pre-loaded with every installed pack and the assessments, dossier filings, and maturity assessment record types. Tool affinity for search_regulatory and detect_framework_conflicts.

Contract-fluent persona for legal review
Owns clause libraries, redline review, and jurisdiction-aware risk language. Pre-loaded with the contract corpus and the GDPR, PDPL, DORA, and NIS2 packs. Tool affinity for search_documents, read_document, and the redline workflow.

Privacy-fluent persona for the data protection officer
Owns DPIAs, RoPA, transfers, breach notifications, and rights requests. Pre-loaded with the PDPL, GDPR, and Cybersecurity Law packs plus the seven privacy-lane record types. Tool affinity for the privacy workflow pack and search_regulatory.

Risk-fluent persona for the second line
Owns risk acceptances, vendor risk, and sub-processor change impact. Pre-loaded with vendor records, risk acceptances, related contracts, and security-relevant packs (ISO 27001, SOC 2, DORA, NIS2). Tool affinity for the vendor and risk workflow pack.

Auditor-fluent persona for the third line
Owns walkthroughs, control evidence sufficiency, and privilege-creep detection. Pre-loaded with the ISO 27001 and SOC 2 packs plus the assessments and audit lineage record types. Tool affinity for the audit workflow pack and the lineage view.

WHAT FORSETI WILL NOT DO
The features ride on top.


Every guardrail is intentional.
Nothing is left to chance.

Every decision is traceable.
Every action leaves a trail.

No Persona will conclude legally without an inline citation naming the framework and the specific requirement.

No Persona will accept an off-topic question into general chat – each persona includes an explicit refusal block and a redirect.

No Persona will approve or certify control evidence on the AI side. Human sign-off is always required.

Forseti will not restrict Persona switching. Permissions govern what users can do, not which Persona they're talking to.
See how the same compliance question is answered differently by each specialist persona, with role-specific reasoning, tools, and grounded citations.


Experience all six Forseti personas using your own compliance scenarios, regulatory packs, and organisational workflows.

Talk with our team about your compliance goals and which Forseti personas best fit your organisation and workflows.
There are six: a Generalist plus five specialists – the Compliance Specialist, the Legal Counsel, the DPO Advisor, the Risk Analyst, and the Auditor's Eye. Every customer gets all six on every tier. The personas are runtime configurations, not separately trained models.
Yes. Persona switching is a one-click action available on every chat surface. The conversation context persists across the switch; only the voice, tool affinity, and citation discipline change. There are no permissions or tier rules that restrict which persona a user can switch to.
Yes, on the Enterprise tier. The tenant-level system prompt override applies to any persona and is recorded with a legal-review attestation in the audit trail. Customers on Light and Standard tiers use the default persona prompts.
Personas shape voice and tool affinity, for example, the Legal Counsel persona is biased toward search_documents and the redline workflow, while the Compliance Specialist persona is biased toward search_regulatory and detect_framework_conflicts. Personas do not change what Forseti is structurally allowed to do; the constraints (no autonomous filing, no delete tools, no AI-side evidence approval, no legal conclusion without grounding) apply to every persona identically.
By design. The avatars are Glitnir-court geometric motifs, no portrait, no photorealistic face, no name styling that implies a real person. An AI that pretends to be a person is harder to govern than one that presents as a tool.