AI Features Policy

Pinsot is a product operated by Dakoli LLC, a California limited liability company. This document explains the rules, rights, procedures, and responsibilities that apply to review-first artificial-intelligence assistance offered through pinsot

Pinsot provides subscription software and technology for nail salon operations in a global market. Salons are independent businesses that provide nail, beauty, wellness, and related services to their clients. Unless a document expressly says otherwise for a specific transaction, the salon—not Pinsot—is responsible for its services, prices, technicians, schedules, employment relationships, commissions, taxes, refunds, client care, licensing, sanitation, and regulatory compliance.

The English version is the controlling version unless applicable law requires otherwise. The Vietnamese translation is provided for meaningful access and convenience and is intended to preserve the same substance.

Depending on workspace configuration and rollout status, Pinsot AI may assist with receptionist conversations, suggested replies, scheduling context, turn explanations or recommendations, reporting insights, translation, and content generation. AI features labeled Preview are not generally available, may be limited to approved customers or markets, and may change or be withdrawn. Planned features are educational descriptions only until explicitly enabled.

AI output is generated from patterns and available context, not human understanding. It may be inaccurate, incomplete, outdated, biased, inconsistent, or unsuitable for a salon's situation. A confident tone does not establish correctness. Users must review the source facts, permissions, audience, tone, policy, and downstream effect before using output.

An authorized person must review AI output before it changes a booking, communicates a material promise, affects a turn, changes a price, influences compensation, resolves a complaint, restricts access, or is added to a lasting record. Pinsot must provide a meaningful handoff or override where the workflow supports consequential action.

AI must not independently make irreversible or legally significant decisions about employment, worker classification, scheduling rights, pay, discipline, termination, financial approval, refunds, safety, legal rights, medical conditions, discrimination, or account termination. The salon remains responsible for decisions within its control and should use qualified professionals where appropriate.

AI output is not legal, accounting, tax, employment, payroll, medical, insurance, or financial advice. It does not diagnose nail, skin, infection, allergy, or other health conditions. Users must not present an AI response as a licensed professional opinion, regulatory determination, or guaranteed business result.

If a client raises a possible health or safety concern, stop any automated flow that could minimize risk and follow the salon's lawful safety process, including referral to an appropriate licensed professional or emergency service when warranted. For legal, tax, employment, or financial questions, preserve the facts and consult a qualified advisor.

An enabled AI workflow may process the user's prompt, minimized salon context, generated output, feedback, safety classifications, source references, approval decision, and audit information. Data use follows the Privacy Policy, salon instructions, applicable DPA, feature notice, and verified provider configuration. We do not promise that a particular AI provider or processing location applies until verified.

Do not enter passwords, authentication codes, API keys, payment credentials, bank details, government identifiers, unnecessary health information, immigration information, private personnel allegations, trade secrets not approved for the workflow, or unrelated client data. Redact names and details when the task can be completed with roles or synthetic examples.

Salons must train authorized users, set access and approval boundaries, monitor output quality, review recurring failures, keep fallback workflows, and disclose automated assistance to clients when required or reasonably expected. A client must have a practical way to reach a human for sensitive, disputed, complex, or unsupported requests.

Users must not misrepresent AI as a human, conceal a material automation decision, fabricate citations, or use generated content to harass, discriminate, deceive, infringe, or evade consent. Translation must be reviewed by a qualified speaker before high-impact or client-facing use because names, numbers, urgency, dialect, and legal meaning may be lost.

If AI output is harmful, unexpectedly discloses information, contradicts source facts, or attempts an unauthorized action, stop using the output, preserve a redacted example, record the feature and approximate time, and contact support@pinsot.com or security@pinsot.com as appropriate. Do not repeat the unsafe action to gather more evidence.

Pinsot may limit or suspend AI access for unsafe inputs, policy violations, abuse, provider restrictions, security risk, or reliability concerns. We may change models, routing, safeguards, limits, or supported use cases. Material feature changes will be reflected in status labels, notices, or policy updates as appropriate.

This document is version 2026.07.20, effective and last updated July 20, 2026. We may update it to reflect product, legal, safety, or operational changes. Material changes will be communicated through an appropriate public or in-product notice when reasonably required. Continued use after an effective change is governed by applicable law and any notice or consent that law requires.

Questions and requests concerning this document may be sent to support@pinsot.com. Contact addresses are centralized for secure routing. Do not email passwords, verification codes, full payment credentials, government identification, health details, or unrelated client records.