Connect your first tool in about 2 minutes. Data syncs within 15 minutes. The signal engine runs every 30 minutes and the compound engine every 12 hours. Most workspaces see their first ranked decision in the Radar within the first hour.

No. Radar+ is built for operators, not analysts. No SQL, no spreadsheets, no BI-tool configuration. You connect tools with OAuth (or upload a CSV) and the signal engine handles normalization, scoring, and ranking.

A fully-populated demo workspace with realistic signals, action plans, a live Ledger, and the full Decision Inbox UI — so you can click through the exact flow before connecting your own tools. No credit card required.

Most teams are operating off of Radar+ within an hour. Connect tools (2 min each via OAuth), wait for the first sync (≈15 min), and decisions start appearing in the Radar as they compound. No implementation project.

Those dashboards only show what happened inside each tool. Radar+ reads all your tools together and finds patterns between them — the churn signal that's hiding across support tickets + failing payments + stalled renewals only exists in the gap between your tools. That's what the compound signal engine catches.

Each signal becomes a ranked decision: a headline, the accounts driving it, a drafted action plan (Slack posts, HubSpot tasks, Gmail drafts, Stripe retries, calendar invites, etc.), and an approve / modify / reject button. Approved decisions execute through your connected tools and are logged to the Ledger for 14 / 30 / 60 / 90-day outcome tracking.

Every signal shows a confidence score (0–100), the data sources behind it, and the reasoning. You can dismiss, snooze, archive, or reject with a reason. Every rejection feeds the learning loop — detectors and recommendations that produce real outcomes float up; patterns that don't quietly fade.

Every 15 minutes for most connected integrations. The signal engine runs every 30 minutes. The compound signal engine runs every 12 hours (or on-demand after a fresh sync). Webhook endpoints exist for real-time events from Stripe, n8n, Zapier, and Make.

The plan lands in your tasks list — each step has the channel (Slack, email, CRM, etc.), the recipient, the message to send, and the deadline. You open your own Slack, Gmail, or CRM and do the work there. Radar+ is read-only and never writes to your connected tools.

No. Radar+ is read-only. We pull data from Stripe, HubSpot, Intercom, etc. to surface signals, and we write back nothing. Every recommended action is yours to take in your own tools.

Every saved decision pins the target metric, its baseline value, and the expected direction. Radar+ automatically checks back at 14, 30, 60, and 90 days and records whether the metric moved as expected. Those outcomes feed a score multiplier on the detector and recommendation so future rankings tighten over time.

Since Radar+ never writes, there is nothing to roll back on our side. If you took an action in your own tool (a Stripe refund, a Gmail send), you reverse it in that tool. The Ledger keeps a record of what we recommended so you can find what you did and undo it cleanly.

TLS at the host layer, AES-256-GCM for any credentials or tokens we store. Row-level security in Postgres isolates every workspace. We never share data between workspaces.

Never. Your data is used exclusively to generate signals and decisions for your workspace. We don't share data between workspaces, sell to third parties, or use customer data to train external models. The AI layer that generates summaries runs on a stateless API — no customer data is retained by the provider.

Yes. Disconnect any integration at any time from Settings; the OAuth tokens are invalidated immediately. You can also delete the entire workspace, which permanently removes all associated data (signals, decisions, ledger, snapshots, credentials).

We're pursuing SOC 2 Type II. Our infrastructure runs on Vercel and Supabase, both SOC 2 Type II certified — every database query, OAuth handshake, and write through a connected tool runs on attested infrastructure today. See the Compliance section for the deeper breakdown.

Radar+ is in active SOC 2 Type II observation with a Big-Four audit firm; the Type I letter is available under NDA. Every layer of the stack we run on is already SOC 2 Type II — Vercel (compute), Supabase (Postgres + auth), Cloudflare (edge), AWS (object storage). Until our own attestation letter is issued, finance teams can rely on the same controls those vendors maintain (encryption, access reviews, change management, incident response).

Every approved decision is a logged control event: who approved it, what was approved, the source signal, the channel it fired through, the returned provider IDs, and the 14/30/60/90-day outcome. The Ledger is exportable as CSV for your auditor and is immutable once written — modifications create a new row instead of editing the old one. For finance-led workflows (refunds, dunning, AR collection, deferred revenue), this gives you a defensible paper trail without spreadsheet reconciliation.

Yes. Every signal, decision, approval, modification, rejection, integration write, and OAuth event is timestamped, attributed to a user, and retained for the life of the workspace (7+ years configurable). Read-side queries (who viewed which signal when) are also logged. The combined trail satisfies the evidence requirements for SOX 404, SOC 2 CC7, GLBA, and most internal-audit playbooks.

Radar+ is not currently HIPAA-eligible — we don't sign BAAs and the platform isn't intended for PHI. If you need a covered-entity workflow, the suggested integration is a non-PHI metric layer (e.g. portal login counts, billing throughput) that lives outside your EHR. Email ab@roradar.com if your use case needs covered-entity data and we'll route to a partner.

You retain full ownership of your data. We act as a processor under GDPR; the DPA is available at /legal. CCPA right-to-delete and right-to-know are handled in-app — disconnect any integration to revoke OAuth tokens immediately, or request workspace deletion to permanently remove all signals, decisions, ledger entries, and snapshots within 30 days. EU data-residency on the database side is on the roadmap (Supabase EU region) — reach out if you need it pre-purchase.

Yes — invite them as a Viewer (free per workspace). Viewers can read every signal, decision, action plan, ledger row, and the full execution log without being able to approve, modify, or reject. The audit trail records their reads as well, so your evidence chain stays intact.

For an active workspace: indefinitely, with no enforced cap. After cancellation: 30-day soft-delete window during which you can restore, then permanent deletion of all data. KPI snapshots and the Ledger are exportable as CSV at any time so you control the offline copy of record.

Yes — annual third-party pentest, results available under NDA. Plus continuous Dependabot + Snyk scanning on the codebase, and Vercel's WAF on every public route. Email ab@roradar.com to request the latest report.

Status page at status.roradar.com (24/7 monitoring, public RCA within 5 business days). For confirmed data incidents affecting your workspace, the contractual notification window is 72 hours from confirmation — typically much faster in practice. Every workspace owner is on a notification list that's separate from marketing.

Stripe, HubSpot, Salesforce, QuickBooks, Xero, Intercom, Zendesk, Slack, Jira, Monday.com, Airtable, Google Analytics, Google Sheets, OneDrive / Excel, Gusto, and CSV upload — 16 live sources. Many more (Pipedrive, Shopify, Linear, Mixpanel, Snowflake, and others) exist as registry entries on the roadmap.

Slack (chat post), HubSpot (task), Gmail (draft), Intercom (message or conversation reply), Zendesk (ticket), Stripe (refund / credit / dunning retry), and Google Calendar (event invite). Unsupported channels are logged with a clear "manual follow-up" state so the flow never silently fails.

You can upload data via CSV, connect via a Zapier / n8n / Make / Airbyte webhook, or request a new integration. Most requested integrations ship within 2–4 weeks.

Read access is the default. Write access is requested per action, per integration — only when you approve a decision that needs it. You always see which scopes an integration requests at connect time.

Starter is $499/mo ($424/mo billed annually). Core is $1,499/mo ($1,274/mo billed annually). No per-seat fees. No usage caps. Price is per workspace.

Yes — the sandbox is free and shows every feature end-to-end. Connect your own tools when you're ready. No credit card required to start.

Investors get free viewer access to any workspace they're invited to. Each portfolio company pays for their own workspace. One VC relationship can monitor 5–25 companies at no additional cost.

Your data remains accessible for 30 days after cancellation. After that, all data is permanently deleted. You can export your signals, Ledger, and KPI snapshots as CSV at any time.