clariBI
clariBI is an AI-powered business intelligence platform that links Stripe, Shopify, Google Analytics and 175+ other tools, then answers cross-source questions in plain English. Dashboards build themselves; no SQL and no data team required.
What is clariBI?
clariBI is a managed business intelligence platform built for companies whose numbers are scattered across a dozen SaaS accounts and who have nobody in-house to join them up. Its stated premise is that a real business question rarely lives inside one tool: revenue sits in Stripe, traffic in Google Analytics, contacts in HubSpot, orders in Shopify. Rather than opening five dashboards, you type the question once and the platform assembles the answer from every source at the same time.
Connection is deliberately undemanding. Native OAuth connectors cover Google Analytics, Google Ads, Meta Ads, Google Sheets, Jira and Confluence; direct connectors reach PostgreSQL, MySQL, SQL Server, Oracle and SQLite; and a catalogue built on the Model Context Protocol adds Stripe, HubSpot, Linear, Notion, Atlassian, PostHog, Klaviyo, Sentry, Vercel and many more. The site advertises 175+ integrations in total, all read-only, plus CSV, Excel and PDF upload for anything left over.
On top of that sits the analytical layer. Questions are asked in ordinary English and come back as charts, written insights and explanations. Dashboards are generated by the system itself and refreshed every morning with the previous day's complete data. A template marketplace offers more than 200 sector-specific starting points. The forecasting module projects a metric 30 days out, running nine methods through a walk-forward backtest, and returns a prediction band together with driver discovery under false-discovery-rate control, plus anomaly and changepoint detection.
Governance is treated as a first-class feature rather than an afterthought: role-based access control with five roles and more than 35 permissions, TOTP multi-factor authentication, OAuth single sign-on with Google, Microsoft, GitHub and LinkedIn, SAML 2.0 for Okta, Azure AD and OneLogin, and an append-only audit table with per-row hashes. SMS-based MFA is deliberately absent, the vendor calling it phishable.
Developers get two surfaces: an MCP server exposing 19 tools with OAuth 2.1 and API keys, and a REST API on the higher tiers, wrapped by official Python and TypeScript SDKs. There is no mobile app and no self-hosted edition.
What it does
- Answer business questions asked in plain English, across every connected source at once
- Join data held in separate tools inside a single analysis, such as Stripe revenue against HubSpot engagement
- Generate dashboards automatically and refresh them each morning with the previous day's complete data
- Project any metric 30 days ahead and surface the drivers behind the movement
- Flag anomalies and structural breaks on any KPI being tracked
- Publish and share reports, including through secure public links that can be revoked
- Drive the whole platform from Claude Desktop, Cursor or ChatGPT through its MCP server
When to use clariBI / When not to
A quick filter to help you decide if clariBI is the right fit.
When to use clariBI
- Founders and small SaaS teams tracking MRR, churn and CAC without hiring a data analyst
- Marketing agencies that need to prove campaign ROI to clients with automated, shareable reports
- E-commerce and retail operators combining Shopify orders, ad spend and customer segments in one view
- Finance and operations managers who rebuild the same weekly report by hand every Monday
- IT and procurement teams under InfoSec review, who need MFA, SAML SSO and exportable audit logs
When not to use clariBI
- Teams that need a mobile app: clariBI is web-only, with no iOS or Android release
- Organisations requiring self-hosting or on-premise deployment, since clariBI is managed SaaS only
- Buyers who must show a SOC 2 or ISO 27001 certificate, as the vendor states plainly that it holds none
- Tight budgets that still need MFA, SSO or an audit trail, all three being excluded from the $99 Starter tier
- Analyst-led teams fluent in SQL, dbt or LookML, for whom the vendor itself points to Mode, Hex or Looker
How to use clariBI
A typical end-to-end flow, from setup to results.
- Create an account and start the 14-day trial; no credit card is requested
- Connect a first source through one-click OAuth, for example Google Analytics, Google Ads, Meta Ads, Jira or Confluence
- Alternatively point clariBI at your own database, choosing PostgreSQL, MySQL, SQL Server, Oracle or SQLite
- Upload a CSV, Excel or PDF file, or ingest from a public URL, for anything that has no connector
- Add further vendors from the MCP catalogue, such as Stripe, HubSpot, Linear, Notion or PostHog, still read-only
- Let the platform generate its first dashboards on its own from what is now connected
- Ask a question in plain English in the conversational panel and read the chart and explanation it returns
- Save the answers you want to keep into a dashboard, then adjust the widgets and the layout
- Start from a marketplace template rather than a blank page, set goals and OKRs, and run a forecast on a key metric
- Invite the team, assign RBAC roles, switch on MFA and SSO, then share by public link or scheduled report
Pros & Cons
Pros
- Prices are published in full, from the free tier to $999 a month, with no mandatory sales call
- A 14-day trial runs without a credit card and drops back to a permanent free plan, data intact
- Very broad integration surface, 175+ vendors, connected read-only through OAuth in a couple of minutes
- Forecasting and driver discovery are included in the base product, not sold as an enterprise add-on
- OAuth SSO ships from the Professional tier; the vendor openly criticises the industry's "SSO tax"
- The audit trail is an append-only table with row hashes, exportable as JSON or CSV on Enterprise
- Unusual candour: comparison pages say when a rival is the better choice, and the vendor states plainly which features it has not built
Cons
- No security or compliance certification at all, neither SOC 2 nor ISO 27001, as the vendor itself acknowledges
- The $99 Starter tier excludes MFA, SSO and the audit trail entirely
- SAML 2.0 and audit-log export are reserved for Enterprise at $999 a month, and the REST API starts at Professional
- No mobile application and no self-hosted option; the product is web-only managed SaaS
- AI credits are capped per tier, from none on Free and Lite to 500 on Starter, and must be topped up beyond that
- The pricing page depends on JavaScript: the tiers never render without it, which also hides them from crawlers
- The publisher is a Czech sole trader rather than an incorporated company, which is worth weighing for continuity
Pricing & Plans
A permanent free plan is available: "Free forever" gives one user three data sources, one gigabyte of storage and auto-generated dashboards, but no AI credits. A 14-day trial adds full AI analysis without a credit card and reverts to the free plan on expiry, with data retained. The cheapest paid entry point is the Lite tier at USD 19.00 per month, or USD 190 per year. Starter costs USD 99 per month, Professional USD 199 and Enterprise USD 999, each discounted by 17% on annual billing.
- 1 user
- 3 data sources
- 1 GB storage
- no AI credits
- 5 file uploads
- auto-generated dashboards
- 1 user
- 3 data sources
- 1 GB storage
- 50 AI credits
- MCP server access included
- 1 user
- 5 data sources
- 2 GB storage
- no AI credits
- manual dashboards
- MCP excluded
- 3 users
- 10 data sources
- 5 GB storage
- 500 AI credits per month
- MCP integrations
- 15 users
- 50 data sources
- 10 GB storage
- 1
- 500 AI credits per month
- RBAC
- scheduled reports
- audit viewer
- 100 users
- 100 data sources
- 100 GB storage
- 5
- 000 AI credits per month
- SAML SSO
- audit export
- audit retention of one year or more
- Annual billing carries a 17% discount on the Lite
- Starter
- Professional and Enterprise tiers
Data, GDPR & hosting
A consolidated view of how clariBI handles your data.
GDPR overview
Implementation is documented rather than merely asserted. The Cookie Policy states compliance with the EU GDPR, the UK GDPR and the ePrivacy Directive, and a consent banner runs before any non-essential cookie. The Privacy Policy of 27 December 2025 names its legal bases article by article, 6(1)(b) contract, 6(1)(a) consent, 6(1)(f) legitimate interests and 6(1)(c) legal obligation, and lists the full set of data-subject rights with a 30-day response at privacy@claribi.com. The publisher is established in the EU; transfers to its US sub-processors rely on the European Commission's Standard Contractual Clauses, and a copy of the safeguards is available on request. Breaches are notified within 72 hours. The named supervisory authority is the Czech ÚOOÚ. No DPO is designated, and no Article 27 representative is needed for an EU-based publisher.
Who owns the data?
Under the Terms of Service you retain ownership of every dataset you upload or process, and clariBI claims no ownership rights over it; the Privacy Policy adds that business data is never sold, rented or traded. In exchange you grant the publisher, Darek Černý, a sole trader registered in Prague under IČO 01378091, a licence to store, analyse, back up and run AI and machine-learning algorithms on that data solely to deliver the service. Four named sub-processors handle it: Google for the Gemini language models, DigitalOcean for hosting, Stripe for payments and SendGrid for transactional email. The software, algorithms and trademarks remain the vendor's exclusive property, and a merger or asset sale may transfer your information to the acquiring entity.
Reuse rights
You may reuse your own data freely, without asking permission. clariBI offers self-service export at any time, in CSV, JSON or Excel, either from the platform or on request to support, and says explicitly that the formats are chosen to keep migration easy. On the Professional and Enterprise tiers the REST API returns dashboards, reports, data sources and usage metrics as JSON over HTTPS, while the MCP server can generate reports straight into an LLM client. Two limits apply. The Terms forbid using the service for competitive analysis or to build a rival product, and you remain responsible for holding the rights to whatever you upload. Separately, the vendor reserves the right to use aggregated, anonymised data to improve its own algorithms.
Data retention & training
Hosting summary
The publisher states that it is based in the European Union and that Czech law governs the contract, with the Czech courts holding exclusive jurisdiction. Hosting and infrastructure are provided by DigitalOcean, described on the Privacy Policy only as "USA/EU"; no specific hosting country is named anywhere on the site. Three further sub-processors sit in the United States: Google for the Gemini models, Stripe for payments and SendGrid for email. Transfers to them rely on the European Commission's Standard Contractual Clauses, an assessment of the destination country's legal framework and additional technical measures, and a copy of those safeguards can be requested at privacy@claribi.com. Data is encrypted with AES-256 at rest and in transit over TLS, and data-source credentials are encrypted separately under a rotating key. Backups are encrypted and stored in geographically distributed locations. Note one internal inconsistency: the Privacy Policy cites TLS 1.3 while the security page says TLS 1.2 or above.
Things to keep in mind
Risks and trade-offs to weigh before adopting clariBI.
- Questions you ask are sent to Google for processing by the Gemini models, so business data leaves the vendor's own infrastructure
- OAuth connections are read-only but can be broad; a source connected carelessly exposes more than the analysis actually needs
- With no external certification, the entire security posture rests on the vendor's own statements
- The publisher is an individual entrepreneur with no corporate structure, which raises the continuity risk compared with an incorporated vendor
- Data is permanently deleted 30 days after cancellation, so a forgotten export becomes an irreversible loss
- Trusting an AI-generated figure without cross-checking it is easy, and there is no analyst in the loop by design to catch a wrong answer
- Handing the analytical reasoning to the tool can quietly erode the team's own data literacy, and a forecast shown with a prediction band is a range, never a promise
Setup & Integrations
Technical difficulty
Low for the intended user. The vendor claims a first data source connected in under five minutes and most connections done in under two, through OAuth rather than shared passwords, with no semantic layer to model beforehand and dashboards generated automatically on first connection. A quick-start guide and a knowledge base cover the first dashboard. Difficulty rises for direct database connections, which need credentials and network access, and for the REST API or MCP server, which assume developer skills. SAML 2.0 on Enterprise requires configuration on the identity-provider side.
Deployment
Integrations
Behind clariBI
Social
Resources
All the official URLs gathered for verification and reference.
Alternatives
Tools that compete with or complement clariBI.
Frequently asked questions
Do I need to know SQL to use clariBI?
Is there a free plan, and what does the cheapest paid tier cost?
Does the free trial require a credit card?
What are AI credits and how far do they go?
What can clariBI connect to?
Can I use clariBI from Claude Desktop, Cursor or ChatGPT?
Is there a mobile app?
Who owns the data, and where is it hosted?
Is clariBI SOC 2 or ISO 27001 certified?
How long is my data kept after I cancel?
Should you pick clariBI?
clariBI occupies a narrow, clearly stated position: business intelligence for companies that have data in a dozen places and nobody employed to join it up. Everything about the product follows from that choice. Connection is one-click and read-only, questions are asked in plain English, dashboards assemble themselves, and forecasting with driver discovery ships in the base product rather than behind an enterprise quote. The pricing is published in full, from a permanent free plan to USD 999 a month, which is rarer in this category than it should be.
What makes the site unusual is its candour. The comparison pages name the situations where Tableau, Looker, Mode or Metabase is the better answer. The security page lists what has not been built — WebAuthn, SCIM, a bug bounty — and states outright that the product holds no compliance certification. That honesty is genuinely useful when evaluating, and it is also the main caveat: an organisation that needs SOC 2 or ISO 27001 on paper will not get it here.
Two further reservations deserve weighing. Security features that many buyers consider baseline — multi-factor authentication, single sign-on, the audit trail — are absent from the USD 99 Starter tier, so the real price of a governed deployment is Professional or above. And the publisher is a Czech sole trader, Darek Černý, rather than an incorporated company, with a domain registered only in June 2025; that says nothing about product quality, but it is a continuity question worth asking before entrusting it with core reporting.
For a small SaaS, agency or e-commerce team that wants cross-source answers this week rather than a data-warehouse project this year, the free plan and the card-free trial make the evaluation cheap to run.
- Choosing a selection results in a full page refresh.
- Opens in a new window.