Praxa Privacy Notice

Scope

This notice explains how Praxa, provided by Nexis Labs, collects, uses, shares, retains, and protects information when you use the Praxa app and services.

Agreeing to the Praxa Terms of Use and acknowledging this notice does not enable Adaptive Context, precise location enrichment, camera, microphone, motion, calendar, contacts, health, memory, or other optional sources. A separate Enable Praxa Context action turns on a limited low-sensitivity set plus matching calendar or location context only when the operating system already allows it; it does not open system permission prompts. Camera, microphone, motion, health, communications, and precise Google location enrichment remain off until separately enabled. One master switch can revoke every Adaptive Context source, and every source remains individually revocable.

Information we process

The information Praxa processes depends on the features you use and the choices you make.

• Account and profile information, such as your account identifier, name, avatar, preferences, and authentication records.
• Conversations, files, feedback, memories you choose to save, and the outputs needed to respond to your requests.
• Tool, connection, automation, approval, and transaction records needed to perform and audit actions you request.
• Device capability and permission status, app diagnostics, crash reports, and usage analytics according to your Settings choices.
• Optional Adaptive Context signals only for sources you separately enable. Camera frames, microphone audio, motion, location, calendar, health, and similar sensitive sources follow the source-specific disclosure shown before or beside their controls.
• Subscription and purchase status. Payment providers process payment credentials; Praxa does not store full card numbers or security codes.

Precise location and Google enrichment

Google location enrichment is not granted by the legal acknowledgement checkbox, the Praxa Context master switch, or the operating-system Location permission. You must enable its separate source control. When enabled, Praxa sends precise coordinates through Praxa's bounded, ephemeral server broker to Google Maps Platform only for a direct nearby-area, travel-time, air-quality, pollen, or related location request.

Air Quality, Pollen, Nearby Places, Isochrone, and Places Aggregate results appear only in current, attributed result cards. The assistant can request the matching card, but Google Maps content is not passed to the language model, voice model, memory graph, or AI training. Provider payloads, photos, request coordinates, and raw Google responses are removed before conversation persistence.

Direct-user Isochrone and count-only Places Aggregate responses are display-only, zero-retention, and ineligible for model or memory use. Elevation remains blocked until verified product-purpose and presentation requirements are implemented. Praxa does not add request coordinates or raw Google responses from this enrichment path to your memory graph, prompts, or AI training data.

You can turn Google location enrichment off at any time in Settings. Turning it off stops future enrichment requests; operating-system location permission is controlled separately on your device.

Praxa includes Google Maps features and content. Use of those features and content is subject to the Google Maps/Google Earth Additional Terms of Service and Google Privacy Policy, linked below this notice.

Health, camera, and voice context

Apple Health data is read on your device only for a request you make after you separately enable Health context and approve the matching Health permission. A small, time-stamped summary may be sent to the model selected for that request, but the Health query, raw samples, response draft, and Health tool output are removed before conversation or tool-ledger persistence. Health data does not enter Adaptive Context rhythms, the neural world model, or memory.

When you separately enable a camera source and ask Praxa to use it, Praxa briefly analyzes a foreground camera frame on your device. Frames, face images, identities, raw landmarks, and raw labels are not uploaded or stored. Only a bounded current-turn cue, such as attention, lower-energy appearance, or a coarse surroundings category, may be sent to the selected model; that cue is removed before conversation and tool-ledger persistence. Praxa does not infer identity, emotion, diagnosis, or sensitive traits from a face.

When you separately enable Voice context and use active voice mode, Praxa may derive bounded speaking-pace and audio-energy counts while the microphone is already active. Praxa does not retain microphone audio or a transcript in this path, and the derived cue is cleared when voice mode ends, consent is revoked, or verification fails. Praxa rechecks Voice context consent during an active session and does not use these cues to infer emotion, identity, diagnosis, or sensitive traits.

How we use information

We use information to provide and personalize requested assistant responses; run tools, automations, and approved actions; maintain memory and context features you enable; secure accounts; prevent abuse; diagnose failures; improve reliability and quality; administer subscriptions; and comply with law.

Praxa does not sell personal information. We do not use a legal acknowledgement as permission to activate optional sensors or context sources.

How information is shared

We disclose information only as needed to operate Praxa, complete a request you direct, protect users and the service, or meet legal obligations.

• Infrastructure, database, authentication, model, monitoring, and communications providers that process data for Praxa.
• Tool and data providers you choose to use, including Google Maps Platform for separately enabled location enrichment.
• Payment and app-store providers for purchases and subscription administration.
• Authorities or other parties when required by law, needed to protect rights and safety, or involved in a merger, financing, acquisition, or sale subject to appropriate safeguards.

Retention and deletion

We retain information for the shortest period reasonably needed for the feature, security, legal, and operational purposes described here. Retention differs by data type. Ephemeral context can expire quickly; account, approval, transaction, and legal receipt records may be retained longer where needed for security, audit, or legal obligations.

You can review or remove supported memories and context sources in Settings, revoke connected services, and request account or data deletion. Some records may remain where law, fraud prevention, dispute resolution, or security requires it.

Your choices

Settings provides one master Praxa Context switch plus individual source controls, including a separate precise Google location enrichment control. Memory, proactive suggestions, permissions, analytics, and diagnostics remain reviewable, and device permissions can also be changed in your operating-system settings.

Depending on where you live, you may have rights to access, correct, delete, restrict, object to, or receive a copy of personal information. Contact us to make a request; we may need to verify your identity.

Security and international processing

We use technical and organizational safeguards designed to protect information, including authentication, access controls, encryption in transit, and source-specific data minimization. No system can guarantee absolute security.

Praxa and its providers may process information in countries other than where you live. Where required, we use recognized transfer safeguards.

Children

Praxa is not directed to children under 13, or the higher minimum age required in a jurisdiction. We do not knowingly collect personal information from children who cannot lawfully use the service.

Changes and contact

We may update this notice as Praxa changes. When a material change requires renewed acknowledgement, Praxa will present the new version before you continue; optional permissions and context consents remain separate.

Questions or privacy requests can be sent to team@nexislabs.io.