Findly
Findly answers natural-language questions about company data by pairing large language models with a proprietary semantic layer. It connects GA4, Facebook Ads, BigQuery and SQL databases, then returns charts, dashboards and scheduled reports via web, Slack, Teams or API.
What is Findly?
Findly is a natural-language analytics and forecasting platform published by Conversion Pattern Inc., a Delaware company, and sold under the Findly brand. Its stated method is to pair large language models with an in-house semantic layer, which the vendor presents as what keeps answers accurate and reliable rather than merely plausible. That layer is described once in a section called Catalog — measures, dimensions, identifiers and metrics — and the modelling step is skipped entirely when the source is Google Analytics.
Documented connectors cover Google Analytics 4, Facebook Ads, Google BigQuery, PostgreSQL and Microsoft SQL Server; the pricing table adds Google Search Console and Shopify Ads, and opens connection to any SQL database on the Enterprise plan. Answers come back as charts, real-time dashboards that can be explored, multi-format exports and report templates. Findings then travel to where teams already work: scheduled reports go out by email, Slack and Microsoft Teams, and the assistant can be called in a channel with @Findly. A REST API completes the picture — a token is created from account settings, documented endpoints cover a health check and conversation creation, and two OpenAPI specifications are published.
Beyond descriptive analysis, Findly offers predictive analytics on key indicators and scenario modelling built on historical data. The use cases it puts forward are sector-specific: commodity trading, energy trading and oil prices, supply chain work on routes, cargo and delivery times, consulting, retail and e-commerce.
The range spans four products: Findly AI for enterprise, Darling for energy trading, Semantica, an AI-ready data layer, and Wayward for e-commerce and marketing. A partner offer adds white-label deployment and an embedded assistant, alongside consulting, enterprise, industry-solution and creator programmes.
The homepage claims more than 8,000 users, SOC 2 Type II and ISO 27001 compliance, and a Google Cloud Marketplace listing. The company came out of the Y Combinator Summer 2022 batch and displays partnerships with Google Cloud, Google for Startups Accelerator AI-First, Mistral and Perplexity for Startups. Three co-founders lead it: Jonathan Aeschlimann (CEO), Pedro Nascimento (CTO) and Lourenco Maciel (CPO), with a team presented as coming from Google, Meta, Microsoft and Twitter. A separate British entity, Findly UK Limited, is registered in London.
What it does
- Ask a question about your data in plain language and get an answer back
- Connect GA4, Facebook Ads, BigQuery, PostgreSQL or Microsoft SQL Server
- Model measures, dimensions, identifiers and metrics in the semantic layer
- Build real-time dashboards and drill into the results
- Schedule recurring reports to email, Slack or Microsoft Teams
- Produce forecasts and run scenario models on historical data
- Query the assistant from Slack or Teams, or open a conversation through the REST API
When to use Findly / When not to
A quick filter to help you decide if Findly is the right fit.
When to use Findly
- Business teams that need answers from company data without writing SQL or queuing behind an analyst
- Data teams that want governed metrics defined once in a semantic layer and reused by everyone
- Commodity and energy traders following price forecasts, cargo movements and market risk
- Supply chain, retail and e-commerce operators tracking routes, delivery times and demand trends
- Consultants and marketing teams turning GA4, Facebook Ads and Search Console data into client-ready analysis in hours
When not to use Findly
- Anyone expecting a mobile app: no iOS or Android client is offered
- Individuals and tight budgets: beyond a free tier capped at Google Analytics, Facebook Ads and three users, the first paid plan is 149 USD per month
- Teams that insist on testing before paying: there is no free trial of the paid tier, and access to Standard is granted on request
- Buyers who need SAML SSO, on-premise deployment or a connection to any SQL database, all held back for the Enterprise plan and its unpublished price
- Procurement teams with strict GDPR requirements: no data processing agreement, no sub-processor list and no hosting country are published
How to use Findly
A typical end-to-end flow, from setup to results.
- Create an account on the Findly dashboard — the free Basic plan requires no payment
- Connect a data source; anything other than Google Analytics requires the Standard or Enterprise plan
- Build the semantic layer in the Catalog section, where a setup assistant automates most of the modelling
- Skip the modelling step altogether if Google Analytics is your only source
- Ask your question in the Assistant section of the dashboard
- Follow the query on the Data requests page, then visualise or export the result
- Add the Slack or Microsoft Teams integration to query the assistant from a channel
- Schedule recurring reports to email, Slack or Teams
- Generate an API token in account settings and pass it as a request header for programmatic access
- Request access to move up to the paid Standard plan, which is not sold self-service
Pros & Cons
Pros
- An explicit semantic layer: metrics are defined before they are queried, not improvised at answer time
- A permanent free plan with unlimited messaging for up to three users
- Warehouse and database connectors (BigQuery, PostgreSQL, Microsoft SQL Server), not only marketing sources
- Insights delivered where teams already are: Slack, Microsoft Teams and email
- A documented public API, with two OpenAPI specifications published
- A written commitment that customer and provider data is never used to train AI models
- Y Combinator Summer 2022 backing, a Google Cloud Marketplace listing and partnerships with Google Cloud, Mistral and Perplexity
Cons
- No GDPR compliance is claimed anywhere; the documentation mentions only a 'GDPR-readiness' being pursued, and no data processing agreement, sub-processor list or hosting country is published
- The homepage states SOC 2 Type II and ISO 27001 compliance while the product documentation FAQ says the company is still pursuing SOC 2 compliance: the two claims contradict each other
- The advertised trust centre publishes no readable content, so neither certification can be checked there
- The 149 USD per month headline is a 70% launch discount off a 497 USD list price, not a stable rate
- No free trial of the paid tier: access to Standard has to be requested rather than bought
- The free plan connects only Google Analytics and Facebook Ads, capped at three users
- The features that decide a rollout — SAML SSO, white-label, any SQL database, on-premise — are Enterprise-only, and the Enterprise price is not published
Pricing & Plans
Findly offers a permanent free plan, Basic, with no time limit and unlimited messaging for up to three users. The lowest paid entry point is the Standard plan at 149.00 USD per month, a figure that reflects a 70% launch discount applied to a 497 USD list price and is shown alongside a 1,788 USD annual billing amount; the discount is presented as a launch offer and should not be treated as a standing rate. The Enterprise plan is quoted on request and its price is not published. No free trial of the paid tiers is offered: access to Standard is granted on request. All prices are displayed in US dollars.
- up to 3 users
- Google Analytics and Facebook Ads connections
- date comparisons
- report generation and export
- report templates
- unlimited messages per workspace
- email support
- everything in Basic
- up to 10 users
- Google Analytics
- Facebook Ads
- Shopify Ads
- Google Search Console and BigQuery connections
- state-of-the-art AI models
- scheduled email reports
- everything in Standard
- unlimited users
- the earlier connectors and more
- connection to any SQL database
- white-label deployment
- embedded assistant
- SAML SSO
- on-premise deployment
Data, GDPR & hosting
A consolidated view of how Findly handles your data.
GDPR overview
Findly makes no GDPR compliance claim. The regulation is named once across the site and the product documentation, in the documentation FAQ, which states that Findly is pursuing SOC 2 compliance and GDPR-readiness, to meet enterprise needs — an objective in progress, not a declared conformity. No Article 27 EU representative is designated, no data processing agreement is published or offered, no sub-processor list exists, and no hosting country or region is disclosed. The publisher, Conversion Pattern Inc., is a Delaware company and the terms are governed by United States law before Delaware courts; the Findly UK Limited entity shown in the footer is not an establishment in the European Union. Data subject rights are nevertheless open in practice: the privacy policy grants access, correction and deletion of personal information, exercised at privacy@findly.ai.
Who owns the data?
Findly's privacy policy leaves personal data with the individual: users have the right to access, edit or delete their personal information from the platform, conversation history can be removed at any time, and requests go to privacy@findly.ai. On business data, the Services page states that your data remains on your server, maintaining the privacy and integrity of your valuable information. The terms take the opposite line on the software itself: the site, the service and their content remain the intellectual property of Findly and its licensors. In short, you keep your data and your account record, while the publisher, Conversion Pattern Inc., keeps the platform.
Reuse rights
Findly's terms reserve intellectual property in the site and the service to Findly and its licensors. What the platform produces, by contrast, is built to travel: results can be visualised, exported in several formats, shared publicly and pushed into scheduled email, Slack or Microsoft Teams reports. On Findly's side, the privacy policy states that customer and provider data is never used to train AI models, and that datasets are not retained, stored or combined for any form of AI training or algorithm improvement. All processing is strictly limited to delivering the requested service functionalities unless the customer authorises more in writing. Data reaches Slack only when the user switches that integration on, and only to the extent the connected service requires. Personal information is not sold to third parties, and Google Analytics is used for website audience measurement.
Data retention & training
Hosting summary
Findly discloses no hosting country and no hosting region. Two product statements bear on the question: the Services page says that customer data remains on the customer's own server, and the documentation FAQ answers the question of private data access with "No, Findly uses a secure and scalable architecture to ensure data security and compliance." An on-premise deployment option appears in the pricing table, reserved for the Enterprise plan. The publisher, Conversion Pattern Inc., is established in the United States (Delaware), with a separate British entity, Findly UK Limited, in London, and the terms are governed by United States law. The domain resolves to an anycast IP address allocated to Amazon, which reflects the website's delivery infrastructure rather than any statement about where product data is stored. In practice, buyers with data residency requirements have nothing published to work from: no region, no country, no sub-processor list and no data processing agreement. Those points have to be obtained directly from the vendor.
Things to keep in mind
Risks and trade-offs to weigh before adopting Findly.
- The 149 USD per month price is a 70% launch discount off a 497 USD list rate: budget for the full price in case the offer ends
- 'Unlimited Messaging' is capped in practice: the terms state that the term "Unlimited" does not imply the capability for infinite or unrestricted use
- The homepage claims SOC 2 Type II and ISO 27001 compliance while the documentation FAQ says SOC 2 compliance is only being pursued: ask for the certificates before relying on either
- No GDPR compliance is claimed, no Article 27 EU representative is designated, no data processing agreement is offered, and no sub-processor list or hosting country is published; the Findly UK Limited entity in the footer is not an EU establishment
- Paid plans are governed by a separate SaaS Agreement that is not public, under United States law and Delaware courts: read it before signing
- Fluent answers are easy to trust without checking. The semantic layer only reflects what was modelled in the Catalog, so an incomplete or wrong metric definition yields a confident figure that is still wrong — reconcile numbers against the source before acting on them
- Maintenance signals are mixed: the footer copyright is frozen at 2024 while the legal documents date from 2025, and no incorporation date is published
Setup & Integrations
Technical difficulty
Difficulty depends entirely on the source. Creating an account is free and immediate, and connecting Google Analytics requires no modelling at all, which is the short path. Any database is heavier: measures, dimensions, identifiers and metrics must be modelled in the Catalog, though a setup assistant automates much of it, and connecting BigQuery, PostgreSQL or Microsoft SQL Server means supplying credentials and network access. Semantic layer guidance is included from the Standard plan and dedicated installation support from Enterprise. Slack and Teams integration is documented step by step. The API route, with token generation and headers, calls for developer skills.
Deployment
Integrations
Behind Findly
Fundraising
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Does Findly access or store private data?
Which platforms does Findly connect to?
Is my data used to train AI models?
Is there a free plan?
How much is the first paid plan, and is there a free trial?
Is there an API?
Can Findly be used inside Slack or Microsoft Teams?
Do I have to configure a semantic layer?
Which security certifications does Findly hold?
Is there a mobile app, and is there a minimum age?
Should you pick Findly?
Findly's bet is straightforward: conversational analytics only holds up if the numbers behind the answers are governed, so the product places a semantic layer between large language models and company data and treats it as the guarantee of accuracy. That choice shows everywhere — measures, dimensions, identifiers and metrics are modelled in a Catalog before anyone asks a question, and the modelling step is only skipped when Google Analytics is the source.
What supports the case: a permanent free plan, connectors that reach past marketing sources into BigQuery, PostgreSQL and Microsoft SQL Server, a documented REST API with published OpenAPI specifications, delivery into Slack, Microsoft Teams and email where teams already work, and a written commitment that customer data never trains a model.
What tempers it: the site never claims GDPR compliance, the only mention being a 'GDPR-readiness' still being pursued; no data processing agreement is published, no sub-processor list exists and no hosting country is disclosed. The advertised trust centre publishes nothing readable, and the homepage's SOC 2 Type II and ISO 27001 claims sit awkwardly beside a documentation FAQ describing SOC 2 as work in progress.
The economics deserve the same caution. The 149 USD per month entry price is a 70% launch discount off 497 USD and is presented as promotional; there is no trial of the paid tier, only a request for access; and the features that decide an enterprise rollout, from SAML SSO to white-label, any SQL database and on-premise, are Enterprise-only at an unpublished price.
Findly is worth the detour for data-literate teams that want governed self-service analytics and can live with United States law and a thin compliance paper trail. Buyers working through a strict GDPR checklist should obtain the missing documents in writing before committing.
- Choosing a selection results in a full page refresh.
- Opens in a new window.