Formula Bot
Formula Bot is an AI data analyst: upload a spreadsheet or a PDF, or connect a live database, then ask questions in plain language and get charts, dashboards, scheduled reports and slides without writing SQL or Python.
What is Formula Bot?
Formula Bot is an AI-driven data analysis platform published by Datasetmatch LLC, a United States company. The principle is straightforward: you upload a file or connect a source, ask a question in everyday language, and the AI returns answers, charts, tables and reports. The home page claims more than one million users. The product is organised around ten modules, namely Chat with Data, AI Agents, Data Context, Scheduled Tasks, PDF to Excel, Embedded Analytics, Data Enrichments, Data Explorer, Presentations and Dashboards, and twelve advertised capabilities covering visualisation, transformation, code transparency, AI agents, text analysis, connectors, scheduled reporting, Data Explorer, embedded analytics, Data Context, Brand Kits, spreadsheets and reports. What sets it apart is code transparency: the Python, SQL or R generated for each request stays visible, verifiable and exportable, so an analysis can be audited rather than trusted blindly. Three AI model tiers are offered, Base for speed, Advanced for most tasks and Expert for complex analysis, running on OpenAI, Google and Anthropic models. Input coverage is broad. Seventeen connectors reach Google Analytics, Google Search Console, Google Ads, Tableau, Power BI, Airtable, Google Sheets, Google Drive, Dropbox, OneDrive, SharePoint, BigQuery, Snowflake, MySQL, PostgreSQL, Supabase and Microsoft SQL Server. Uploaded files may be xlsx, xls, xlsm, csv, tsv, native or scanned PDF with OCR, docx, pptx, txt, mp3 or wav for transcription, and video. Data Explorer adds curated public datasets on government, economics, health and weather that can be analysed directly. A family of free satellite tools surrounds the platform: Excel and Google Sheets formula generators, SQL, VBA, Google Apps Script and regex generators, PDF-to-Excel and bank-statement converters, sentiment analysis, homework help and calculators. Each chat runs in its own isolated, dedicated sandbox, row-level security is available on database connections, hosting is on AWS in the United States, and encryption uses TLS in transit with AES-256 at rest. Enterprise customers can embed a white-label iframe inside their own product, with per-user row-level security. The site is published in English and Spanish, and the home page states that the tool can be questioned in any language.
What it does
- Ask a question in plain language about an uploaded file or a connected database and get an answer, a chart or a table
- Generate charts, KPIs, sparklines and dimension filters automatically, with the AI choosing the representation
- Clean, merge, reshape, filter and reformat messy datasets without writing formulas
- Convert PDFs, scanned tables and bank statements into structured spreadsheets
- Build interactive dashboards from a template and schedule daily, weekly or monthly refreshes delivered automatically
- Generate Excel formulas, SQL queries, VBA, Google Apps Script and regular expressions
- Export to Excel, CSV, PDF, PNG or SVG, turn a finished analysis into slides, or embed the analytics chat in your own product
When to use Formula Bot / When not to
A quick filter to help you decide if Formula Bot is the right fit.
When to use Formula Bot
- Business teams that sit on data but have no dedicated analyst to interrogate it, since every question is asked in plain language and needs no SQL, Python or any programming language
- Analysts and BI professionals who want to stop hand-building dashboards and recurring reports, defining an analysis once and rescheduling it daily, weekly or monthly
- Excel and Google Sheets users wrestling with messy files that need cleaning, merging, reshaping and reformatting before they can be read
- Finance and back-office staff who spend their time turning PDFs, scanned tables and bank statements into usable spreadsheets
- SaaS vendors that want to put a white-label analytics chat in front of their own customers through an embedded iframe
When not to use Formula Bot
- Teams handling sensitive personal data: the privacy policy explicitly asks users not to upload Social Security numbers, health information or biometric data
- Engineering teams that need to plug an analytics engine into their own pipeline, as there is no public API, only a white-label iframe embed
- Organisations bound by European data residency, because hosting is announced in the United States only and no EU region is offered
- High-volume users hoping to stay on the free plan, which is capped at 10 messages and 25 AI actions per month, 5 MB per file and five uploads in total, with irreversible deletion after 60 days of inactivity
- Under-13s, who are excluded from the service outright, and 13- to 18-year-olds without parental consent
How to use Formula Bot
A typical end-to-end flow, from setup to results.
- Create an account with an email address and password or a Google sign-in at analytics.formulabot.com; no credit card is required and email verification uses a six-digit code
- A question typed on the home page carries over into the free workspace once you have signed up, and the site promises a first result in 30 seconds
- Add your data: upload a file, connect a source, or combine several sources in the same chat
- Authorise Google, Airtable or Power BI connectors through OAuth; database connectors ask for credentials, an open firewall on your side and sufficient permissions
- Ask your question in plain language in the chat
- Read the chart, table or answer, and open the generated Python, SQL or R code if you want to check the calculation
- Export to Excel, CSV or PDF for tables and PNG, SVG or PDF for charts, or turn the result into a dashboard or a set of slides
- Build a dashboard: pick a template (Executive, Terminal, Glass or Editorial), let the AI assemble it, then edit each element by clicking on it
- Schedule a recurring task: save the analysis, then choose the frequency (daily, weekly or monthly), the day, the time and the time zone
- Go further by declaring Data Context for connected SQL databases, creating reusable agents and Brand Kits, installing the Excel and Google Sheets add-ons from the Starter plan up, enabling TOTP multi-factor authentication, or pasting the iframe snippet to embed the chat in your own application
Pros & Cons
Pros
- No technical skill required: everything runs through natural language, with no SQL, Python or programming language to learn
- The generated code stays visible and exportable, so the analysis can be audited instead of being taken on trust as a black box
- Broad input coverage: 17 connectors across analytics, BI, databases, warehouses and cloud files, plus a wide range of formats including scanned PDF with OCR and audio
- One continuous chain from question to chart to dashboard to slides to scheduled report, with a white-label embed for SaaS vendors on top
- A permanent free plan with no credit card, and a refund policy expressed in figures: 30 days and fewer than 50 messages or actions consumed
- A clear, repeated commitment never to train models on customer data, backed by a subprocessor list published by name
- Real account-level security: row-level security on database connections, TOTP multi-factor authentication, TLS and AES-256 encryption, and an isolated sandbox per chat
Cons
- No public API is documented, so integration is limited to the iframe embed
- The free plan is very tight (10 messages and 25 AI actions per month, 5 MB per file, five uploads in total) and everything inactive is irreversibly deleted after 60 days
- Hosting is in the United States only, with no EU data residency option
- Governance is thin: no postal address published, one generic email address covering legal, GDPR and support, no Article 27 EU representative, no DPO and no customer DPA
- No certification of its own: the SOC 2 Type II cited belongs to AWS, the host, not to Formula Bot
- Real cost is hard to forecast: unused messages do not roll over between periods, Max and Enterprise are priced per seat, and a single operation consumes between 1 and 10 AI actions
- No native mobile application, and support runs by email only, with founder chat reserved for Enterprise
Pricing & Plans
Formula Bot offers a permanent free plan at 0 USD per month, limited to 10 messages, 25 AI actions, one file per chat and five uploads in total, with a 5 MB ceiling per file. The lowest paid entry point is the Starter plan at 18 USD per month, or 13 USD per month when billed annually at 151 USD per year. Above it, Max costs 29 USD per seat per month (20 USD per seat per month billed annually, 244 USD per year) and Enterprise 149 USD per seat per month (99 USD per seat per month billed annually, 1,188 USD per year); the home page presents the annual commitment as saving around 30%. Payments are processed by Stripe by card and may be cancelled at any time. A refund may be requested within 30 calendar days of the last payment provided fewer than 50 messages, requests or actions have been consumed since, and also in cases of duplicate billing, billing error or persistent platform failure. Custom plans are quoted on request, and Enterprise embeds include three embed users on the first embed, then 10 USD per month per additional user.
- 10 messages per month
- 25 AI actions per month
- one file per chat and five uploads in total
- 5 MB per file
- base AI model
- 5-minute execution timeout
- two-factor authentication
- community support
- 250 messages per month
- 2
- 500 AI actions
- 20 files per chat
- 200 MB
- premium models with extended memory
- 4 CPU and 2 GB
- 10-minute timeout
- unlimited messages
- 5
- 000 AI actions per user
- 10 automated dashboards
- 6 CPU and 4 GB
- 20-minute timeout
- 50 tasks including 10 scheduled
- 500 MB and 25 files per chat
- 15
- 000 AI actions per user
- unlimited shared agents and Brand Kits
- 20 automated dashboards
- 8 CPU and 8 GB
- 30-minute timeout
- 1 GB and 60 files per chat
- 50 connected sources
- Every paid plan includes 5 reference files per agent
- and a custom plan can be quoted through the contact form or a booked call
- with negotiated AI action volumes
- dashboards
- connectors and embed controls
Data, GDPR & hosting
A consolidated view of how Formula Bot handles your data.
GDPR overview
Compliance is claimed in a single line on the security page: Formula Bot states that it complies with the General Data Protection Regulation for EU users. Beyond that sentence the record is thin. No Article 27 EU representative and no data protection officer is named in the privacy policy, the terms or the security documentation, and no postal address is published anywhere. The company is based in the United States and says information may be processed and stored there and wherever its service providers operate; no standard contractual clauses, Data Privacy Framework certification or adequacy decision is mentioned. Access, rectification, deletion, export and portability rights are exercised by writing to hello@formulabot.com, with possible identity verification; in embedded deployments the request may go through the Customer as controller. No customer-facing DPA is published. Privacy policy in force since 19 August 2026, terms last updated 25 June 2026.
Who owns the data?
Section 5.1 of the terms states that, as between the parties, you keep ownership of your content: uploaded files, prompts and outputs, or the Customer for Customer Data. Formula Bot takes only a limited, worldwide, non-exclusive, royalty-free licence to host, store, reproduce, process, transmit and display that content, solely to deliver the service. Embedded deployments shift the roles: the Customer is generally the controller and Formula Bot the processor, and it may reach account identifiers, usage logs, chat histories, outputs and files depending on its configuration. Feedback and suggestions are the exception, usable by the publisher without restriction or compensation. The home page sums it up: "Your data is YOUR data".
Reuse rights
Because ownership stays with the user, the files, prompts, answers, charts and exports produced in Formula Bot can be reused freely, without asking permission. On the publisher's side, data serves to provide, operate and improve the service, authenticate users, keep accounts secure, process payments, run the connectors, deliver support, fight abuse and meet legal obligations. Prompts, messages, uploaded files and context are passed to third-party AI providers, namely OpenAI, Google and Anthropic; voice dictation is sent to OpenAI for transcription and the audio itself is not kept by Formula Bot. Two exclusions are stated plainly: personal information is neither sold nor shared for cross-context behavioural advertising, and data pulled through Google APIs is never sold and never used for advertising, its use being confined to the connectors requested, in line with Google's Limited Use policy. Google Analytics measures site traffic, and cookies cover operation, preferences and audience measurement. In a merger, acquisition, financing or asset sale, data may be transferred with the business.
Data retention & training
Hosting summary
Data is hosted on Amazon Web Services, in United States data centres, with geographic redundancy announced for high availability. Traffic is encrypted with TLS in transit and data at rest with AES-256 on RDS. The application layer is deployed through Render and the public site runs on Bubble.io, while code execution takes place in an E2B sandbox isolated per chat. Isolation also relies on row-level security in the database, a dedicated storage folder per user, and connection credentials encrypted through an AWS encryption service. Security headers are active on the site: HSTS, XSS protection and MIME sniffing blocking. Formula Bot presents AWS as covering 143 security standards and certifications, including SOC 2, ISO 27001 and PCI DSS, which are the host's credentials rather than the publisher's own. No EU hosting region is offered, which rules the platform out wherever European data residency is a requirement.
Where Formula Bot works
Country-level availability.
Not available in
Things to keep in mind
Risks and trade-offs to weigh before adopting Formula Bot.
- Publisher identity is thin: no postal address appears anywhere on the site, the only identifiable entity is Datasetmatch LLC, and a single generic address, hello@formulabot.com, handles legal, GDPR and support requests alike
- GDPR compliance is claimed without an Article 27 representative, a DPO, a customer DPA or any mention of standard contractual clauses, while data is stored in the United States only, with no EU residency option
- The security credentials on display belong to the host: the SOC 2 Type II and the 143 standards and certifications cited are AWS's, not Formula Bot's
- Three third-party AI providers, OpenAI, Google and Anthropic, receive prompts and uploaded files, and the no-training commitment on their side is conditional, applying "where available and applicable to our account configuration"
- On the free plan, anything inactive for 60 days is permanently and irreversibly deleted; database credentials, by contrast, are never removed automatically and stay until the user or the administrating Customer deletes them
- In embedded deployments, the Customer can reach the end user's chat histories, files and usage logs
- The publisher itself acknowledges that AI outputs can be inaccurate and that reviewing them is the user's responsibility, which is a real over-reliance risk when the answer arrives as a finished chart; the site's structured data also declares a 2022 founding while the domain has been registered since 2015, and the home page testimonials cannot be traced to any public profile
Setup & Integrations
Technical difficulty
Getting started is easy: sign up with an email address or a Google account, with no card, and the site promises a first result in 30 seconds, since uploading a file is enough. No SQL, Python or programming skill is required, and Google, Airtable and Power BI connectors take a few OAuth clicks. Difficulty rises with databases, which need credentials, a firewall opening on your side and sufficient permissions. Data Context and row-level security assume knowledge of the schema, the Excel and Google Sheets add-ons install separately, and embedding means pasting an iframe snippet, a front-end task.
Deployment
Integrations
Supported languages
Behind Formula Bot
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
What is Formula Bot?
Do I need to know SQL or Python?
Which file formats can I upload?
Which data sources can I connect?
Is there a free plan?
Is my data used to train AI models?
Where is my data stored and how long is it kept?
Can I export my analyses?
Is there an API?
How accurate are the results?
Should you pick Formula Bot?
Formula Bot is a mature tool that covers the whole chain, from a question typed in plain language to a chart, a dashboard, a set of slides and a report delivered on schedule. Its distinctive strength is code transparency: the Python, SQL or R generated for each request stays visible and exportable, which turns an AI answer into something a careful analyst can actually audit. The commitment never to train models on customer data is repeated consistently, and documentary transparency is decent overall, with subprocessors named, retention quantified and a refund policy expressed in figures rather than intentions. Governance is the weak side. No postal address is published anywhere, a single generic email address serves legal, GDPR and support requests, no Article 27 EU representative or data protection officer is named, no customer DPA is offered, and the certifications cited belong to AWS, the host, not to the publisher. Hosting is exclusively in the United States, with no EU region: for any organisation bound by European data residency, that alone is disqualifying. The free plan is genuinely useful for evaluating the product but very constrained, with ten messages a month and irreversible deletion after sixty days of inactivity. The paid entry point at 18 USD per month is reasonable for a solo user; per-seat pricing on Max and Enterprise changes the scale for a team, and the AI action metering makes intensive use hard to budget in advance. In short, a strong fit for business teams and analysts who want to stop building dashboards by hand and are comfortable with United States hosting, and a poor fit for regulated environments, European data residency requirements, or engineering teams looking for a programmatic API.
- Choosing a selection results in a full page refresh.
- Opens in a new window.