Terms of Service

Controller: KGPX Serviços Administrativos LTDA (trade name TalkWithData). CNPJ: 23.543.381/0001-50. Address: Rua Gustavo Ambrust, 464 — Campinas/SP CEP 13092-106, Brazil. Data Protection Officer (encarregado): Guilherme Favaron. Contact: [email protected].

Questions about these Terms: [email protected].

You must be at least 16 years old to use the Service. By registering, you represent that you meet this requirement and have the legal capacity to enter into these Terms.

If you use the Service on behalf of an organization, you represent that you have authority to bind that organization to these Terms.

You are responsible for maintaining the security of your account credentials. Do not share your password with third parties.

You must provide accurate and complete information during registration and keep it updated.

We reserve the right to suspend or terminate accounts that violate these Terms or engage in fraudulent, abusive, or illegal activity.

You agree NOT to:

- Upload data that you do not have the right to access or analyze.

- Attempt to access, probe, or disrupt the Service infrastructure, including servers, databases, or networks.

- Use the Service to process illegal, harmful, or unauthorized data.

- Reverse engineer, decompile, or attempt to extract source code from the Service.

- Use the Service to build a competing product.

- Abuse the AI or query system with excessive or malicious requests.

- Upload malware, viruses, or other malicious code.

- Publish an artifact carrying data you do not have the right to disclose. Publishing makes that version readable by anyone holding the link, and what to publish is your decision.

You retain all rights to the data you upload to the Service. We do not claim ownership of your data.

You are solely responsible for the data you upload, including ensuring you have the necessary rights, consents, and permissions to process it through the Service.

The Service uses third-party AI providers as described in the Privacy Policy. In summary: on upload we may send schema and the rows of your file — or, for a document (PDF, TXT, MD), the text of the document and its file name, under the same context-window rule applied to the model that writes the description, with a fixed ceiling of 128k tokens when that model's window cannot be established — to generate a description and suggested questions; when you ask a question we send schema, description, the current question, and a bounded excerpt of the recent conversation in that chat (at most 6 prior exchanges, without their result rows) to plan SQL — the schema in that call carries a compact profile of your data (the values of low-cardinality columns, date ranges, a few example values), excluding columns that look like personal identifiers; if that SQL fails to execute we may send it back with the engine's error message for correction; when interpreting results we may send the question, SQL, and the rows your query returned. Rows are sent whole up to what the AI model in use can hold — at most 60% of that model's context window, or a fixed ceiling of 128k tokens when that window cannot be established — and when they do not all fit, the answer tells you so. Queries against uploaded data are executed locally. For connected databases, the owner decides whether the rows a query returns may be sent to the AI provider for interpretation: an account-wide switch in Settings and a per-source switch on each database's card, both on by default, and rows travel only while both are on; with either off, the turn delivers the query and the table with a factual count and no AI reading.

We employ reasonable measures to protect your data, including encryption of connection strings at rest and read-only access controls. However, no system is perfectly secure.

We are not liable for unauthorized access to your data resulting from your own credential compromise or misconfiguration.

The Service uses artificial intelligence to generate SQL queries and natural language responses based on your data.

AI-generated content may contain errors, hallucinations, or inaccuracies. You should always verify critical results before making decisions based on AI output.

We do not guarantee the accuracy, completeness, or reliability of AI-generated responses.

We strive to maintain high availability but do not guarantee uninterrupted service. We may perform maintenance, updates, or changes that temporarily affect availability.

We are not liable for downtime, data loss, or service interruptions caused by factors beyond our control.

The Service may offer Free and paid plans. Features and limits are described on the pricing page and may change.

Plan limits are enforced. Your plan's monthly question allowance applies to your account: once you reach it, further questions are refused until the next monthly cycle, and you are notified in the product as you approach it. We may grant an exemption from that allowance to test accounts we authorize. Anti-abuse rate limits apply to every account, exemption or not.

Paid billing (including Stripe checkout) is enabled only when we activate billing features. Until then, paid plan CTAs may be disabled.

We do not currently provide team features such as row-level security (RLS), SSO, or organizational audit logs unless separately documented as shipped.

To the maximum extent permitted by law, we are not liable for indirect, incidental, special, consequential, or punitive damages, or any loss of profits, data, or business opportunities arising from your use of the Service.

Our total liability for any claim relating to the Service is limited to the greater of (a) the amount you paid us in the 12 months before the claim or (b) USD 100.

Nothing in these Terms limits rights that cannot be waived under applicable consumer protection law.

You may delete your account at any time from Settings. Upon deletion, we remove associated account data as described in the Privacy Policy.

We may suspend or terminate access for Terms violations or risk to the Service or other users.

These Terms are governed by the laws of the Federative Republic of Brazil, without regard to conflict-of-law rules.

Exclusive venue for disputes arising from these Terms is the courts of Campinas, State of São Paulo, Brazil, unless mandatory consumer law requires otherwise.

These Terms are provided in English and Portuguese. The version displayed in the Service language you used when you accepted these Terms is the version you accepted.

If the English and Portuguese versions conflict, the Portuguese (Brazil) version prevails.

We may update these Terms from time to time. Material changes will be communicated via the Service or email. Continued use after the effective date constitutes acceptance of the updated Terms, except where applicable law requires a new affirmative acceptance.

Contact: [email protected].