Everything you wanted to ask

Fronterio is The AI Adoption Platform. It helps companies assess their AI readiness, prioritise the initiatives worth doing, drive real adoption across every employee, govern their agents, prove the ROI to leadership, and stay EU AI Act compliant, all on one platform, across every AI tool they use.

Yes. Visit /demo to drop straight into a live dashboard populated with a synthetic 200-person HR-tech company's data. No signup, no email, no credit card. Click around, talk to the AI Consultant, view governance, compliance, board pack, all production components rendering real (synthetic) data. The demo session resets daily, so you can poke without consequences.

Mid-market and enterprise organisations that want AI to actually pay off: AI leads, CIOs, compliance officers, executives, and the employees using AI every day. Every dashboard adapts to the person looking at it.

A lot. Every free account gets a full AI Readiness Assessment with peer benchmarks, the EU AI Act compliance baseline, 10 AI Consultant messages and up to 3 agents — plus the Adoption Engine for your whole team: a personal AI dashboard, daily AI tips, a prompt library, skills challenges, and Fronti, your AI coach (5 Help Desk answers, 5 micro-lesson picks and 5 challenges each month). No credit card, free forever — not a trial.

About 15 minutes. You get an instant score across five dimensions (Strategy, Technology, People, Process, and Culture), plus peer benchmarks and a prioritised action list. Free, no credit card.

Most teams are up and running in a day: run an assessment, register your first agents, switch on governance. Reaching full adoption (metrics flowing, compliance framework in place, the whole team on board) typically takes 30 to 60 days, and the built-in 30-day onboarding flow walks you through it.

Yes. Pro comes with a 30-day free trial. You add a card to start it but aren't charged during the trial, and you can cancel any time before day 31. On day 31 it converts to paid unless you've cancelled. Don't want to enter a card yet? The Free plan never expires and gives you a full assessment, a peer-benchmarked report, 10 AI Consultant messages, and up to 3 agents, enough to evaluate the platform end to end first.

Only the people managing AI strategy need a Pro seat (10 seats included). The Adoption Engine, included in every paid plan with no per-seat charge, gives every employee their own AI dashboard, tips, learning paths and help desk.

Minutes. The Day-One flow lets you pick a quick win for your role, whether a one-page opportunity brief, a clean runbook, or a starter AI use policy, and the platform drafts a real, ready-to-use deliverable in about a minute. You review and edit it, then keep it. It's included on every plan; Free comes with 5 AI drafts a month, and Pro and above are unlimited. No prompt-writing, no setup, no credit card.

A board-grade “State of AI at your company” report: your readiness score by dimension, how you compare to anonymised peers, your company archetype, and what leadership should do next. You can share it through a private, revocable link — it's designed to forward to your CEO or board. Free forever, no credit card.

Yes. Every plan includes assessment invites: Free includes 3, Pro and Business include 5, and Enterprise includes 100. Colleagues answer without needing an account, your organisation's score becomes the mean across all respondents, and the perception gap view shows where leadership and the team see things differently.

Yes. The free AI policy starter tool at fronterio.com/ai-policy-generator builds a nine-section starter policy from a few questions about your organisation — no account needed. You get the policy on screen and as a PDF by email. Paid plans add an AI-tailored policy generator that adapts the policy to your assessment results and governance setup.

Yes. The public prompt library at fronterio.com/ai-prompts has over 40 field-tested prompts across six departments — free to browse, no account needed. When you sign up, you can save any prompt straight into your workspace and share it with your team.

Fronterio is one continuous Operating Loop with four rooms. Design: assess readiness, prioritise initiatives, set strategy (Canvas, Roadmap, OKRs, Board Pack, Workshops, Frontier Firm Designer). Adopt: the Adoption Engine — everyday AI use for every employee. Govern: every agent registered and risk-classified in one accountable estate, EU AI Act compliance built in. Prove: adoption metrics, ROI, board-ready reports. The AI Consultant and Partner Consultant Network sit alongside.

It knows your organisation. The AI Consultant is grounded in your assessment scores, agents, metrics, compliance posture, role, and industry, so its advice is specific, not generic. It adapts its tone to your role, flags compliance implications as they come up, can launch step-by-step Implementation Guides, and hands off to your partner's human consultants when a question goes beyond it.

A context-aware AI engine built on leading frontier models and run entirely server-side in the EU region. The point isn't the model, it's the grounding: Fronterio's AI works from your assessment scores, agents, metrics, and compliance posture, so its answers are specific to your organisation, not generic. We're not tied to a single model, and personal data is never sent to any model API, only aggregated, anonymised, org-level context.

Through the Change Navigator, the platform's change-management layer. It generates your rollout comms kit (leadership announcement, manager talking points with an honest job-impact FAQ, and a town-hall deck) grounded in your readiness assessment, runs an anonymous monthly culture pulse with AI-categorised barrier analysis, activates an AI-champions network with a ready-made playbook, trains managers with a dedicated Leading Change lesson track, and tracks every initiative through an awareness-to-reinforcement change journey. When sentiment slips, the platform recommends the fix that matches the cause: communication, initiative discovery, or skills. Included from the Pro plan.

Three ways. The Initiative Radar suggests 3–5 personalized initiatives every week, grounded in your readiness assessment, your existing registry, and anonymous patterns from companies in your industry. Accept one and it lands in your registry ready for prioritisation. The quarterly Initiative Pulse anonymously asks every employee what eats their time and clusters the answers into promotable candidates. And the public initiative library carries 67+ proven templates across 15 industries. Every plan gets the radar; Free plans see the top suggestion weekly.

Yes. The Governed AI Tool Directory is a curated catalog built for European deployers. Every tool carries three compliance facts on the card: where the vendor can process your data (EU, EU-region-optional, non-EU, or self-hosted), whether a Data Processing Agreement is available, and a plain-language EU AI Act note for deployers. An EU-friendly filter makes the compliant shortlist in one click, and when the Shadow AI Detector finds an unsanctioned tool on an employee laptop, the finding links straight to its sanctioned tier and EU-friendly alternatives.

A curated catalogue of 40+ AI tools and agents, each with compliance notes, risk hints, and one-click adoption into your governance workflow. Pro and up get recommendations tailored to your assessment scores and the tools you already run.

Enterprise visibility into the AI nobody told you about — built on ingestion, not endpoint scanning. Fronterio correlates signals you already have — Microsoft Purview, DLP and network exports, and your declared tool inventory — anonymises user identifiers, and proposes what it finds straight into governance, so unsanctioned tools become registered, risk-classified, and owned.

Yes (Pro and up). The Teams Agent brings the AI Consultant, prompt library, daily tips, challenges, and progress tracking into Microsoft Teams. Your people chat with it, search prompts with @Fronterio, and get proactive nudges without leaving the tool they already work in.

A complete bottom-up adoption programme: Fronti AI coach in Microsoft Teams + help desk, 500+ Copilot scenario lessons grouped into 7 role-based learning tracks (Executive / Strategy / Technology / Finance / Operations / Compliance / General), per-app adoption deltas (Outlook / Teams / Word / Excel / PowerPoint), monthly employee confidence pulse, AI tip feed, skills challenges, leaderboard, prompt library, achievements, champion network, and a 5-step onboarding wizard. Included in every paid plan (Pro, Business, Enterprise) for your whole company, not per seat.

Structured, step-by-step workflows for specific AI scenarios, e.g. "Deploy a Customer Support Agent" or "Set up FRIA for High-Risk AI". Each step adapts to your context, recommends tools appropriate to your role, and can generate deliverables like policies, RFPs, or training plans on demand.

Yes — that's the point. Beyond the live dashboards, Fronterio generates leadership artifacts on a cadence: a quarterly board pack (readiness, adoption, governance and ROI, narrated, with a due date so it's ready before the meeting), a monthly AI adoption report (outcome-first, co-branded for partners), and a yearly renewal ROI dossier. Each is a real PDF designed to be forwarded, not a dashboard someone has to open.

No. Governance and compliance are built in, but Fronterio is the AI Adoption OS. It spans the full loop: assess readiness, design your AI-first operating model, drive employee adoption, govern every tool and agent, and prove ROI to leadership. Compliance is the floor, not the whole product.

A connected strategy layer for leaders, execs, and boards. Five AI-drafted primitives (Strategy Canvas, Strategic Roadmap, Async Workshops, Strategic OKRs, and the Executive Board Pack), each one reading the last. The Canvas drafts your ambition and top moves; the Roadmap turns them into a 12-month plan; Workshops align the exec team asynchronously; OKRs cascade objectives into department key results; the Board Pack composes it all into the one-scroll view you take to the board each quarter. No consultants, no stale slide decks: your strategy lives in the platform and stays in sync with your data.

1) Strategy Canvas (/dashboard/design/strategy) is a Playing-to-Win cascade drafted by the AI: Ambition, Where-to-Play, Top 3 Moves, 90-Day Commitments. 2) Strategic Roadmap (/dashboard/design/strategy/roadmap) is a 12-month quarterly plan generated from the ratified canvas, cascading into tasks and initiatives. 3) Async Workshops (/dashboard/design/workshops) are six opinionated exec templates (Assessment Debrief, Initiative Prioritisation, Archetype Alignment, Roadmap Build, Board Pre-Read, AI Rollout & Change) with AI clustering and AI commitment docs. 4) Strategic OKRs (/dashboard/design/strategy/okrs) are a company → department → key-result cascade with weekly AI refresh from business metrics. 5) Executive Board Pack (/dashboard/design/board) is a real-time one-scroll exec view with an AI-written quarterly narrative bound to live data.

You can, and many companies do: expect €500K–2M, six months to deliverables, and a slide deck that starts drifting the moment it's signed. Fronterio's AI Strategy OS ships the same artifact types (canvas, roadmap, OKRs, board pack) at 1% of the cost, drafted by the AI in under 15 seconds, editable by your leadership team with a sign-off ledger, kept in sync with your live platform data (assessment scores, agents, metrics, compliance posture), and refreshed every quarter without a new engagement. The same goes for your AI-first operating model: the Frontier Firm Designer produces the designed target state in minutes and cascades it straight into initiatives, agents, and obligations. The strategy belongs to you, not to the firm's template library.

Free tier: read-only archetype-default Canvas preview. Pro (€299/mo): AI-drafted editable Canvas + ratification + Strategic Roadmap + Strategic OKR Cascade + async Workshop participation. Business (€699/mo): unlocks Workshop hosting and the full owner-only Board Pack (maturity benchmarks + risk heatmap + AI-written quarterly narrative), plus co-branded external Board share tokens (tokenised read-only links for non-members, optional partner branding for reseller customers). Enterprise: unlimited workshop participants, custom workshop templates.

Yes, from the Business plan (and Enterprise). Generate a tokenised share link and email it to board members. They get a read-only view identical to the in-app page, no login required. The token respects revocation (disable it and the link 404s), is scoped to a single org, and excludes PII. The Board Pack itself starts on Business and is owner-only; owners can also just view it in-app without sharing externally.

Because the two-hour slot where the whole exec team is free at once is a myth. Async Workshop Mode spreads capture, voting, and clustering over 2–7 days, so people contribute over coffee, at lunch, on the train. The AI clusters the input into themes, dot-voting surfaces consensus, and at close the AI writes the commitments doc (who does what by when) and turns those commitments into tasks automatically. (Live 1-on-1 consultant video sessions still exist separately for when you need real-time conversation; async is built for the multi-person alignment problem that live meetings fail at.)

Yes. That's exactly what AI Vendor Resilience does (Enterprise). For each AI workload you nominate a primary vendor plus one to three fallbacks, in priority order. Fronterio auto-discovers what you already use from your Microsoft, Google, and Anthropic connectors, drafts a capability-gap analysis and switchover runbook for each fallback in seconds, and lets you run a failover drill that proves the fallback is genuinely ready (credentials valid, capabilities matched, compliance equivalent, guardrails transferable) without deploying anything. The drill result lands in a tamper-proof log you can show a regulator, and the Board Pack inserts a one-paragraph resilience update every quarter so your board sees it before an outage, not after.

Yes. The curated vendor catalog tags every vendor with data region (US, EU, multi, or EU-sovereign), so EU-sovereign alternatives like Mistral Le Chat Enterprise and Aleph Alpha surface first when you're picking a fallback for a US-based primary. The composite resilience score explicitly weighs EU-sovereign coverage as one of four sub-scores. EU AI Act Article 26(5) asks deployers to monitor operation and maintain continuity plans. A pre-provisioned EU-sovereign fallback per high-risk workload is exactly the operational-continuity control auditors look for.

Generic AI memory is conversational: a chatbot remembering what you said last time. The Fronterio Customer Memory Engine is structural: every signal your organisation generates (assessments, agent lifecycle events, policy publishes, initiative ships, guide completions, decisions, experiments) feeds a nightly consolidation brain that distils durable learned facts across 12 canonical types (preferences, patterns, risk signals, aspirations, commitments, capability gaps, etc.). The facts power recommendations that surface on four surfaces: the AI Consultant prompt (top 5), your dashboard home card (top 3), your weekly digest email (high-priority only), and a dedicated Mission Control page. It's the same durable facts everywhere, ordered by priority. A generic chatbot forgets the context next session; Fronterio remembers your organisation, not just a user.

By design, your raw data never leaves your tenant. Signals are stripped of identifying detail at the boundary: identifiers arrive only as one-way hashes, free text is never accepted, and metadata is limited to a fixed, non-personal set. Your organisation's memory is readable only by your own members. The cross-customer insights (patterns like "organisations like yours usually ship policies in this order") are computed on data that carries no organisation identity and only ever surfaces an aggregate once at least 15 organisations are in it, so your company is never identifiable in a benchmark. And one switch removes you from the exchange entirely.

Yes. Every organisation has a cohort-contribution setting (on by default). Turn it off in Settings and you're excluded both ways: your patterns stop feeding peer benchmarks, and you stop receiving cohort-derived insights on Mission Control and in recommendations. Everything else keeps working: your own learned facts, the nightly consolidation, and your dashboard recommendations are unaffected. The exchange is deliberately two-way, so the trade is always fair: contribute to the network or don't, but either way your own memory keeps working.

A term Microsoft introduced in the 2025 Work Trend Index for AI-first organizations structured around hybrid human-agent teams. It replaces the traditional static org chart with a work chart: per-function AI maturity stages, agent-to-human ratios, and fluid team structures. McKinsey, BCG and Harvard (Eccles, 2025) all converge on similar frameworks.

Stage 1 (AI-Assisted): every person in the function uses AI as a tool. Stage 2 (Agents as Colleagues): agents work alongside humans on specific tasks with oversight. Stage 3 (Agents Run Processes): agents run bounded processes autonomously while humans set direction and handle exceptions. Critically, different functions in the same company operate at different stages at the same time.

Usually executives or the AI Lead, often in collaboration with a Fronterio partner consultant. You map your functions (or pull them from Microsoft 365), set a target stage and agent-to-human ratio per function, and Fronterio generates the narrative and the initiative roadmap. Four minutes to a first designed target state.

When you’re happy with the design, click Generate Function Initiatives. Fronterio creates proposed initiatives in the registry, proposed agents in governance (pre-classified for EU AI Act risk), playbook milestones with auto-generated tasks, and, where stage transitions warrant it, updates the right deployer obligations (Article 4 literacy, Article 14 oversight, Article 27 FRIA). Every generated item is tagged back to the function and the function initiative that spawned it.

Pro (€299/mo) unlocks the design workspace: map your functions, set target stages, visualize the work chart. Business (€699/mo) adds AI-generated function maps, per-function narratives, initiative cascade, and Microsoft Graph sync. Free lets you view a designed state but not edit it.

No. If your tenant admin is happy to grant Directory.Read.All, we pull your real org structure in seconds. If not, Fronterio proposes a realistic function map from your industry, company size, and existing signals from the platform (agents you’ve registered, initiatives you’ve filed, departments you’ve tagged). Both paths end at the same workspace.

Stage transitions trigger concrete deployer obligations. Moving a function to Stage 2 surfaces Article 4 AI literacy requirements. Stage 2 with any high-risk agent surfaces Article 14 human oversight. Stage 3 surfaces Article 27 FRIA. The Designer becomes the forward-looking compliance planner: you see your regulatory surface grow before you ship the agents, not after.

Most customers are deployers: you use AI tools rather than build them. Fronterio focuses on deployer obligations (Articles 4, 14, 26, 27, 50, 73), which is where most of the practical work lives. Provider obligations are covered only where relevant.

Yes. Every EU business needs the baseline, so it's free forever: the compliance dashboard, the 8-obligation tracker, 1 FRIA, AI literacy tracking for 10 employees, a 30-day audit log, the risk-classification wizard with its Article 5 prohibited-practice detector, and manual policy editing. No credit card. Pro adds the automation on top: the nightly Autopilot, AI-drafted policies in 8 languages, the Article 73 incident workflow, weekly post-market monitoring reports, and full audit history with PDF export.

When you register an agent, a structured questionnaire maps its purpose and capabilities to the four EU AI Act categories: unacceptable, high, limited, minimal. High-risk agents trigger additional required fields: conformity assessment, human oversight plan, transparency requirements, technical documentation links.

Before you register an AI agent, a deterministic pattern detector scans the name, purpose, and description for 8 prohibited-practice signatures under Article 5: social scoring, vulnerability exploitation, real-time public biometrics, emotion recognition in workplace/education, untargeted facial scraping, predictive policing by profiling, subliminal manipulation, and biometric categorisation by sensitive attributes. Matches surface a hard warning in the wizard before you can save. Included on every plan.

Yes. A guided wizard (Article 27) walks you through purpose, affected groups, fundamental-rights impact, mitigations, and oversight. The AI Consultant can draft the full FRIA based on your agent data. Export as PDF for regulators.

The AI Literacy tracker records which staff have completed training, when, and the training source. Link it to learning paths in the Adoption Engine for end-to-end evidence, from curriculum to individual completion certificates.

A structured checklist of your Article 26 duties: human oversight, data-input relevance, operational monitoring, log retention, incident reporting, transparency disclosure. Each one has a status, an owner, evidence links, and a due date, and the whole thing exports audit-ready.

Every night at 05:00 UTC, a deterministic engine auto-verifies 6 of your 8 deployer obligations from platform data: AI literacy (Art 4), human oversight (Art 14), operational monitoring (Art 26(5)), log retention (Art 26(6)), FRIA (Art 27), and transparency disclosure (Art 50). The state machine is forward-only: it never downgrades your manual progress.

When you classify an incident as serious, Fronterio auto-computes the 48-hour market-surveillance authority deadline. An hourly cron warns your admin team at t-13 days, t-7 days, t-48 hours, and breach. The platform ships with a seeded directory of competent authorities for all 30 EEA member states. Pro only.

Article 72 requires providers + deployers of high-risk AI to monitor performance after deployment. Fronterio generates a weekly report per high-risk agent covering usage volume, incident rate, human-override rate, complaint signal, and drift alert, drafted as Markdown/PDF and ready for regulator review. Pro only.

The moment you connect M365 Copilot, Google Gemini, Anthropic Claude, or Copilot Studio, Fronterio auto-drafts the Article 50 transparency disclosure in your org's default language (EN/DA/SV/NO/FI/DE/NL/FR), advances the transparency obligation to in_progress, and logs the inference-location + transfer-mechanism (SCCs, EU-US DPF, Microsoft EU Data Boundary) to your audit log.

If your teams build their own high-risk AI systems, you become a provider under Article 16 and face 13 additional obligations (technical documentation, instructions for use, conformity assessment, CE marking, EU database registration, post-market monitoring plan, serious incident reporting, risk management system, data governance, accuracy/robustness/cybersecurity). Enterprise includes the provider obligations suite with auto-generated Annex IV technical documentation from your agent version snapshot.

Yes. The agent registration wizard includes a model-type selector: Agent (default), GPAI-downstream (you use GPT-4, Claude, Gemini as a base), GPAI-provider (you train + release a foundation model). Choose GPAI-downstream and the platform captures the upstream provider documentation URL, tracks it against Article 53 obligations, and adjusts the compliance posture. GPAI providers get an advisory linking to enterprise support, and we onboard those customers with a specialised conversation.

Yes, append-only by design. Entries can only be added, never edited or deleted, and that's enforced at the database layer, not just in the app. Every material action (agent registration, approval, change, policy update, compliance sign-off, incident) is recorded with the user, the timestamp, and the details.

Yes. The compliance dashboard, risk-classification summary, FRIA records, deployer-obligations status, and audit digest all export to branded PDFs (your logo, company name, generation date, clean A4 layout) ready to hand to a regulator or a board.

ISO/IEC 42001:2023 is the first international standard for AI Management Systems. It requires 38 Annex A controls across 9 domains (policies, organization, resources, impact assessment, lifecycle, data, interested parties, use of AI, third-party), plus management system clauses 4–10. Fronterio's Certification Readiness Pack (Business+) auto-maps your existing governance data to all 38 controls, generates a Statement of Applicability in one click, bundles a complete audit pack, and provides built-in management review and internal audit workflows.

The EU AI Act is a regulation: it mandates specific obligations for deployers and providers with legal penalties. ISO 42001 is a voluntary management system standard: it provides a governance framework with certification by accredited bodies. They overlap significantly: ISO 42001 controls A.5 (impact assessment), A.9.3 (human oversight), A.9.4 (risk management) map directly to EU AI Act Articles 27, 14, and 9. If you've already done EU AI Act compliance on Fronterio, your ISO 42001 readiness score starts at 40–60% on day one.

OAuth 2.0 via Entra ID with read-only Graph permissions. Fronterio pulls Copilot usage per user per app, gives you a cross-tenant adoption view, and integrates natively with Copilot Studio: discovery, governance, deployment, telemetry, and an MCP governance bridge.

Yes. The Google Workspace connector (OAuth, Admin SDK Reports API) pulls Gemini audit logs and per-user usage reports. Connect Microsoft and Google side by side for cross-platform comparison, overlap analysis, and combined licence-waste insight.

Yes. The Anthropic connector uses an Admin API key to pull per-workspace-member token usage by model. Microsoft, Google, and Anthropic can all be connected at once for a unified cross-vendor adoption dashboard.

Yes. Fronterio's MCP server is spec-compliant, so n8n workflows connect with zero custom code via either the HTTP Request node (works in every n8n version) or n8n's built-in MCP Client node (v1.45+). Three downloadable workflow templates demonstrate the highest-value patterns: pre-execution guardrails via validate_action (Article 14 evidence), Slack /incident capture for Article 73 deadline tracking, and CI deploy compliance gates. See fronterio.com/integrations/n8n.

Langflow and Dify appear in Fronterio's governed AI tool directory: agents and workflows your teams build there can be registered, risk-classified, and governed like every other part of your AI estate. They are not deployment targets — your teams keep building where they already build, and Fronterio keeps the governance record.

Yes. Register agents built on OpenAI, Azure OpenAI, Ollama, or any other custom or self-hosted model in the governance registry like any other part of your AI estate: each gets an owner, an EU AI Act risk classification, an approval workflow, and an immutable audit trail. Fronterio reads estate data to keep that picture current — it never runs the models itself.

Four (Enterprise): Azure AI Foundry, Microsoft Copilot Studio, Claude Managed Agents, and a custom webhook — one-click deploy from the governance UI after a 7-check pre-deploy compliance gate. Deployment is deliberately a small part of the platform, though: Fronterio's center of gravity is reading and governing your whole AI estate — seeing, correlating, and evidencing what runs across your platforms — not deploying to it.

An SSE endpoint at /api/mcp/sse that external systems (CI/CD, orchestrators, Copilot Studio, LangGraph) can connect to via the Model Context Protocol. They pull approved configs and guardrails, validate actions at runtime, and read the governance state Fronterio maintains — and can report telemetry and incidents back into the registry.

Yes (Enterprise). A one-way sync pushes initiatives, agent registry, compliance status, reports, playbook deliverables, and the weekly audit digest into your SharePoint site. App-only Entra ID authentication: one tenant-admin consent, no per-user login.

Yes. Per-organisation API keys (scoped: read:agents, read:governance, write:telemetry, full) unlock endpoints for agents, assessments, compliance, metrics, shadow-detector findings, tasks, and team. All calls are rate-limited and audit-logged.

Yes. Supabase PostgreSQL in Frankfurt (eu-central-1), Vercel serverless execution in Stockholm/Paris. No customer data leaves the EU. GDPR-compliant with DPA available on request.

Yes. You get data-residency guarantees, a DPA on request, full data export (JSON) and deletion, cookie consent, a published privacy policy, and a documented sub-processor list. Personal data is never sent to the AI, only aggregated scores, anonymised metrics, and org-level context ever leave for processing.

Every table is protected by row-level security scoped to each organisation, enforced at the database layer rather than in application code, so isolation holds even if app logic has a bug. Partners see only their own customers, system admins run on their own access path, and customer users see only their own organisation. Cross-tenant leakage is impossible by design.

Yes, from the Business plan. SAML SSO via any IdP (Entra ID, Okta, Google Workspace, etc.), SCIM for user provisioning, optional MFA enforcement, and SSO-enforced login. It's self-serve: org owners configure it themselves in Settings, no support ticket.

Org owners can export a complete JSON bundle of their organisation's data from Settings at any time. Deletion is one click and permanently removes everything org-scoped (agents, assessments, metrics, conversations, tasks, uploads) within 30 days.

Deployment credentials, OAuth tokens, and tool secrets are encrypted at rest with AES-256-GCM in a dedicated vault. They're never logged, never sent to the AI, and never written to the audit log, and weekly health checks confirm they're still valid.

Because EU AI Act compliance is legal, not commercial. A Free user who registers a high-risk agent is legally required to complete a FRIA, track oversight, and retain logs. Blocking them behind a paywall would make us complicit in their non-compliance. Our upgrade lever is automation + scale, not legal baseline. Competitors charge EUR 10K+/year for this; we give it away.

If you subscribe to Pro or Business and complete our structured 30-day onboarding flow, four board-ready PDFs land in your inbox by day 31: AI Estate Snapshot, First FRIA + Compliance Posture, Policy Pack (3 ratified policies), and a Board-Ready Status Report. If any of the four isn't delivered, we credit your next billing cycle: one month for monthly plans, one quarter for annual. The flow is ~4 hours of work across 30 days; we email reminders along the way. Email hello@fronterio.com to claim.

When you hit any of: a 2nd FRIA, a 2nd agent, an 11th employee in literacy tracking, needing audit history past 30 days, wanting PDF exports for board or regulator, your first serious incident (Article 73 48-hour clock), connecting M365 Copilot / Google Gemini / Anthropic Claude (for auto-wired transparency disclosure), or wanting the nightly auto-evidence engine to run.

When you need the connected estate: AI running on many platforms that leadership wants visible in one accountable registry, estate integrations reading agents and usage from your stack, provider obligations under Article 16 (when your teams build their own high-risk AI), AI Vendor Resilience, and the Brain API — or when you exceed 50 users / 150 agents.

The automation engine, on top of everything in Free: the full EU AI Act compliance suite, the AI Consultant (300 messages/user/month), Implementation Guides with "Generate This For Me", the Adoption Engine for every employee, the Copilot / Gemini / Claude adoption tracker, business metrics and ROI, weekly digests, industry benchmarks, board-ready PDFs, tasks, and Slack notifications. Up to 10 users and 50 agents. Business (€699/month) scales that to 50 users, 150 agents, and 500 messages/user/month.

The connected estate. Enterprise connects Fronterio to everything your company runs AI on: estate integrations that read agents and usage from your platforms, provider obligations (Article 16) for teams building their own high-risk AI, AI Vendor Resilience, the Brain API, Shadow AI visibility from ingested signals, SharePoint sync, SSO/SCIM, unlimited users and agents, a dedicated CSM, custom integrations, higher message limits, and an SLA. Pricing is custom.

Yes. The Adoption Engine is included in every paid plan (Pro, Business, and Enterprise) for your whole company, with no per-seat charge. Employees don't count against the Pro seat limit.

Each Pro user gets 300 AI Consultant messages a month. At 80% you get a quiet heads-up; at 100% you can wait for the monthly reset or buy a €49 pack of 100 extra messages. Enterprise negotiates higher limits.

Yes. Pro Annual is €239/month (billed €2,868/year), ~20% off monthly. Proration is handled automatically by Stripe on upgrades. Downgrades take effect at the end of the current billing period.

Cancel any time from Settings → Billing. You keep access through the end of the period you've paid for. No claw-back, no cancellation fee. We'll ask why you're leaving. It's optional, and only so we can improve.

Reseller (default): Fronterio bills the customer, partner earns 20% lifetime commission on attributed MRR. White-label (premium option): partner bills the customer at wholesale, keeps the margin. Both modes share the same product, dashboard, consultant network, and branding capabilities.

Apply on /partners. System admin reviews within a few business days. On approval you get a partner account, branded dashboard, referral link, onboarding wizard (practice, profile, branding, first client), and access to the marketing toolkit and report generator.

Commission is paid monthly on net revenue (after Stripe fees) for every active customer attributed to you. Structure is configurable per partner: flat (e.g. 20%) or tiered with thresholds and an optional quarterly new-customer kicker. Once calculated, rates are immutable per period.

Optionally, yes. The Partner Sales Incentive Programme pays one-time bonuses at milestones (e.g. €100/signup at 10 customers, €150 at 25, €200 at 50, €250 at 100). Bonuses go to the partner company; internal distribution is at their discretion. Configured per partner.

You register your own consultants. When a customer's AI Consultant escalates, they see your consultant profiles, not a public marketplace. Assigned consultants receive a briefing packet (assessment, conversation summary, metrics) and optionally get a seat inside the customer's org as an embedded expert.

Yes. Custom domain, logo, primary and accent colours, favicon, OG image, AI persona name, email sender domain, and the "Powered by Fronterio" toggle. PDFs carry your branding. In reseller mode you can co-brand; in white-label mode Fronterio branding can be fully hidden.

Yes. Sign up free at /for-employees with your work email and you get a personal AI dashboard, daily tips, learning paths, the shared prompt pool, challenges, a starter Help Desk quota, achievements, and streaks. Free forever, just for you.

Free: personal dashboard, global tip pool, learning modules, public prompts, challenges, 5 Help Desk messages/month, skills score. Needs admin to enable the Adoption Engine: company-specific prompts, champion network, department scorecard, manager insights, digests, Teams integration.

Aggregates only, by default: your employer sees department-level adoption rates, challenge participation, and prompt sharing. Your individual tip interactions, Help Desk conversations, and learning progress stay private unless you choose to share them. And your personal content is never used to train any AI.

Yes (Pro+ orgs with the Adoption Engine). Chat with the AI Consultant, search the prompt library via @Fronterio, get daily tips, check your progress and leaderboard, and receive achievement notifications, all without leaving Teams.