Audacity Advisory Corp Subprocessors
Audacity Advisory Corp, operating as AI Reserve, provides technology, strategic, and advisory services and runs an AI platform that engages subprocessors for infrastructure, model inference, and other client services.
- Subprocessors
- 57
- Locations
- 90
- Categories
- 11
- Last scanned
| Name | Category | Purpose | Location | Data processed |
|---|---|---|---|---|
| Google Cloud Platform | Cloud Infrastructure | Cloud hosting, compute, databases, object storage, secret management, and logging for all AI Reserve services |
 United States
|
Personal Data in the course of providing platform services |
| Google Firebase Authentication | Cloud Infrastructure | Identity and sign-in for the AI Reserve portal (web application) |
United States
|
Identity and sign-in data |
| Auth0 by Okta | Authentication | OIDC identity brokering for keyless desktop-application sign-in (e.g. Claude Desktop) |
United States
|
Sign-in and identity data |
| Resend | Transactional email delivery for the AI Reserve portal (onboarding invitations, password resets, account and allocation notifications, and internal operational reports) |
United States
|
Recipient names, email addresses, and notification content — never prompt or response content | |
| IPinfo | Data Enrichment | IP geolocation and anonymizer screening for the machine API |
United States
|
Request IP addresses only — never request content |
| GDPRLocal Ltd. / Instant EU GDPR Representative Ltd | Privacy | Article 27 EU/UK representative (GDPR Article 27 and UK GDPR Article 27). Receives privacy requests submitted through its portal on AI Reserve's behalf |
 Ireland
 United Kingdom
|
Requester identity and request content only; never prompt or response content |
| Slack Technologies (Salesforce) | Communications | Slack Connect shared channels for customer communications, created from the AI Reserve admin console |
United States
|
Customer admin email address, shared channel name, and messages participants post in that shared channel — including prompts addressed to the @aireserve mention agent and model completions the agent posts back; operational alert payloads may reference customer organization names |
| Microsoft Corporation DPA | Cloud Infrastructure | Microsoft Teams customer channels and Microsoft Entra ID sign-in |
United States
|
Customer contacts' names and work email addresses, channel name, messages and files shared in Teams channels; Microsoft Entra ID tokens carrying tenant id, object id, and sign-in email address. No prompts, responses, or other platform content transits Teams. |
| Daytona Platforms Inc. DPA | Cloud Infrastructure | Isolated, single-use workspaces for the customer Slack agent (@aireserve); persistent workspace volumes and non-ephemeral Linux VM sandboxes for the provider-neutral compute runtime (not enabled, listed ahead of enablement) | United States (New York company sandboxes run in Daytona's United States region) | Request text and bounded Slack thread context, execution-scoped credential for AI Reserve's inference gateway, and optionally authorized repository contents and a short-lived read-only single-repository token; for persistent runtime: snapshot ID, compute limits, network policy, commands, environment values, execution input/output, files, and persistent workspace volume contents |
| Sail Research / Sailboxes | Cloud Infrastructure | Stateful workspaces and elastic virtual-machine sandboxes for the provider-neutral compute runtime; enabled in pre-production (UAT) since November 7, 2026; not yet enabled in production | United States (us-west us-east) enabled in pre-production only production pending acceptance pas | Selected public OCI image, compute limits and network policy, command argument vectors, per-command environment values, execution input and output, files read as artifacts, and contents placed in tenant-scoped persistent workspace volumes |
| Runpod Inc. DPA | Cloud Infrastructure | Customer-configured GPU compute — on-demand GPU pods and persistent network volumes provisioned from the AI Reserve console | United States (company) pods volumes run in the RunPod Secure Cloud data center the organizatio | Container image reference, environment variables, start-up settings, GPU type and data center, SSH public keys, container-registry credentials for private images, and whatever the organization stores on network volumes. Prompts and responses sent through the AI Reserve gateway do not transit RunPod. |
| Apollo.io (ZenLeads Inc.) DPA PP | Sales | Business-contact and company enrichment for the optional chat personas (contact-search and company-team tools) and for the company profile page's team-member and technology-stack sync | United States (EU-U.S. Data Privacy Framework participant Standard Contractual Clauses as fallback | Identifiers a user asks a persona to look up — a person's name and company domain, a LinkedIn profile URL, or an email address — and, on company profile pages, the company's domain; returns business and personal email addresses, phone numbers, title, and location, or company team members and technology stack |
| Exa Labs, Inc. (Exa) DPA PP | Search | Web search for the portal chat assistant. When a chat request calls for current information from the web, the platform sends the search query text to Exa's search API, which returns public web results |
United States
|
Search query text — which can contain terms the user supplied — and nothing else: no other conversation content, no documents, no account data, and no user or organization identifiers |
| Composio (Sampark, Inc. d/b/a Composio) DPA PP | Integration | Connector platform for the portal's optional app connectors. Brokers sign-in and authorization with third-party applications and executes connector tool calls when the user attaches a connected application to a chat conversation | United States (Delaware company San Francisco) | Opaque platform user id, arguments the model generates (which can include content from the user's prompt), and application responses (messages, calendar entries, files, records, or issues from the user's own account). No other conversation content, documents, or account data is sent to Composio. |
| OpenAI, L.L.C. | AI | Model inference (GPT model family; embeddings; image generation) |
United States
|
Request content processed to deliver inference |
| Anthropic, PBC | AI | Model inference (Claude model family, direct API) |
United States
|
Request content processed to deliver inference |
| Amazon Web Services (Bedrock) | Cloud Infrastructure | Model inference for -bedrock and -bedrock-us routed models (Anthropic Claude, Meta Llama, Mistral, DeepSeek serverless) — model publishers do not receive request data | Global (AWS commercial regions via Bedrock global cross-region inference requested from us-east-1) | Request content processed to deliver inference |
| Google Cloud (Vertex AI) | Cloud Infrastructure | Model inference for -vertex routed models (Anthropic Claude partner models, the Gemini line and Model Garden MaaS partner models — xAI, Z.ai, DeepSeek, Meta, OpenAI open weights) — the model publisher does not receive request data | United States (us-east5 Gemini some partner models on Google's global endpoint) | Request content processed to deliver inference |
| Microsoft (Azure AI Foundry) | Cloud Infrastructure | Model inference for -azure routed models (Azure OpenAI deployments of the GPT line, and Foundry Models sold directly by Azure — Anthropic Claude, xAI Grok, DeepSeek, Moonshot Kimi, Meta Llama, Mistral, OpenAI open weights) — Microsoft operates the model as our processor; the model publisher does not receive request data | United States (East US 2) | Request content processed to deliver inference |
| Google (Gemini API) | AI | Model inference (Gemini model family; image generation), paid tier. Also indexes documents saved to the web document library for chat search | United States (processing may occur globally) | Model inference prompts and outputs; saved documents uploaded into a Gemini File Search store scoped to the organization, where Google chunks, embeds, and stores content with no expiry |
| xAI Corp. | AI | Model inference (Grok model family); text-to-speech and customer-created custom voices — organization-gated, off by default for every organization |
United States
|
Prompts and completions; uploaded reference audio for custom voices, with resulting voice model stored by xAI until deleted |
| ElevenLabs Inc. | AI | Text-to-speech (eleven-v4, eleven-v3, eleven-multilingual-v2, eleven-flash-v2-5). Receives the text submitted for synthesis and returns audio | United States (default storage location per ElevenLabs processing may also occur by its internation | Text submitted for synthesis and returned audio; retains both in the account's request history by default |
| Inworld AI Inc. | AI | TTS synthesis (inworld-tts-2, inworld-tts-2-flash), STT transcription (inworld-stt-1), and voice cloning via the /inworld/* HTTP passthrough; additionally inworld-realtime speech-to-speech sessions | United States (registered office processing region pending verification) | Microphone audio, session transcripts, LLM turn context (DeepSeek V4.1 Flash via Inworld's realtime API), and generated audio. Session context is ephemeral and not persisted across sessions. |
| Tako Inc. | AI | Live data and web search (tako-search, tako-search-deep on POST /v1/search). Receives only the search query text and returns knowledge cards and web results |
United States
|
Search query text only |
| Cloudflare, Inc. | Cloud Infrastructure | Decision and extraction models on Workers AI (clef, clef-flash on POST /v1/decisions; clairic on POST /v1/extractions). Receives the request state, typed questions or fields and any embedded images and returns per-answer probabilities or extracted values | Cloudflare global network (United States operator) | Request state, typed questions or fields, and any embedded images |
| Perplexity AI, Inc. | AI | Model inference with web search grounding (Sonar model family); decision models (pplx-decider-v1.2, pplx-decider-flash-v1.2 on POST /v1/decisions) receiving the request state, typed questions and any images embedded in the state, returning probabilities |
United States
|
Request state, typed questions, and any images embedded in the state |
| Together Computer, Inc. (Together AI) | AI | Hosted inference of open-weight models (e.g. Qwen, DeepSeek V4 Pro, Kimi K2.7, MiniMax, GPT-OSS-20B) |
United States
|
Prompts and completions for hosted inference |
| Fireworks AI, Inc. | AI | Hosted inference of open-weight models (e.g. GLM, Qwen, Kimi K2.6, DeepSeek V4 Flash, GPT-OSS-120B) |
United States
|
Prompts and completions for hosted inference |
| DeepInfra, Inc. ✓ | AI | Hosted inference of open-weight and partner models (NVIDIA Nemotron 3 family, DeepSeek V4 Flash, Qwen 3.8 Max) | United States (Delaware corporation HQ Palo Alto CA US data centers plus one Canadian site — Toro | Prompts and completions for hosted inference |
| Novita AI ✓ | AI | Hosted inference of open-weight and partner models (DeepSeek V4.1 Flash, DeepSeek V4 Flash until October 9, 2026, Qwen 3.8 Max) | United States (San Francisco HQ processing regions vary) | Prompts and completions for hosted inference |
| Baseten, Inc. ✓ | AI | Hosted inference of open-weight models (DeepSeek V4 Flash family, GPT-OSS 120B) | United States (San Francisco HQ serving region for our account — see data handling page) | Prompts and completions for hosted inference |
| Sail Research (sailresearch.com) | AI | Hosted inference of open-weight serving legs (api.sailresearch.com): the -sail suffixed variants of Kimi K3, GLM 5.3 / 5.3 Flash, DeepSeek V4.1 Flash / V4 Pro / V4 Flash and Gemma 4 31B, plus the Sail-only Gemma 4 12B and Gemma 4 31B NVFP4 entries | Not fixed — Sail states that by default it uses service providers in multiple regions so customer d | Inference prompts and outputs processed only ephemerally in memory and not persistently stored by Sail or its subprocessors |
| Inworld — Theai, Inc. dba Inworld AI (inworld.ai) ✓ | AI | Two surfaces on one platform account: (1) Hosted inference of open-weight serving legs (api.inworld.ai/v1); (2) The /inworld/* voice passthrough: text submitted for speech synthesis, uploaded audio for transcription and reference audio for instant voice cloning, plus the generated audio Inworld returns | United States — api.inworld.ai is Inworld's US deployment (inputs outputs processed stored i | Prompts and completions in transit; text submitted for speech synthesis, uploaded audio for transcription and reference audio, and generated audio |
| Alibaba Cloud (Model Studio) | Cloud Infrastructure | Model inference (first-party Qwen models — Qwen 3.8 Max, direct Model Studio API) | United States (Virginia) for Qwen chat models (dashscope-us endpoint) except qwen-3-coder-next whi | Request content processed to deliver inference; request data stored in the selected region |
| Tencent Cloud (TokenHub) | Cloud Infrastructure | Model inference (first-party Tencent Hunyuan models and explicit Tencent reseller pins for Xiaomi MiMo, Moonshot Kimi, DeepSeek, Z.ai GLM and MiniMax via Tencent Cloud's TokenHub gateway) | Singapore (Singapore site — tokenhub-intl published resource-scheduling scope "Global" since Novem | Customer Content and Service Generated Content retained up to 30 days unless Zero Data Retention is activated |
| MuleRouter (mulerouter.ai — CarrotHub / SmartStudio ecosystem) | AI | Hosted video generation for the Wan video model families (Wan 3.0 / W3.0, and Wan 2.7 image-to-video), served without upstream content moderation | Singapore — confirmed in writing by the Alibaba Cloud International partner team (August 25 2026) | Video prompt and any customer-supplied media inputs — source/keyframe images, reference images, and reference/driving audio or video, passed by URL or inline |
| WaveSpeedAI PTE. LTD. (wavespeed.ai) | AI | Hosted video extension and prompt-driven image editing for the Wan 2.7 family | Singapore-incorporated operator (WaveSpeedAI PTE. LTD. per its published privacy policy) serving r | Customer-supplied video clip, prompt, optional driving audio, and one to nine customer-supplied reference images, all passed by URL |
| Video Rebirth Limited (BACH by Video Rebirth, bach.art) | AI | Hosted video generation with the BACH bach-1.0 engine: text-to-video, image-to-video, and reference-to-video | Singapore headquarters per the operator's own press releases website terms (Singapore governing | Generation prompt and any customer-supplied first-frame, last-frame, or reference images by public URL; hosts the generated video at a result link |
| Venice.ai (venice.ai) | AI | Hosted video generation for the Wan 3.0 family under Venice's own model ids: text-to-video, image-to-video, and reference-to-video, served without upstream content moderation | United States (the Service "is hosted in the United States" per Venice's privacy policy Wyoming gov | Generation prompt and any customer-supplied source or reference images (by URL or inline); returns the generated video to the platform, which re-hosts it in its own transient storage (24-hour signed link) |
| monid.ai | Data Platform | Data-endpoint marketplace behind /v1/monid (discover, inspect, run) — brokers third-party data endpoints (web scrapers, contact/company enrichment, search) that AI Reserve API callers execute with their own AI Reserve key | Not published pending verification | Natural-language discovery query and the endpoint inputs the caller supplies for a run; forwards run inputs to the third-party provider the caller selected. Never receives chat, completion, or document content. |
| TypeSafe AI | AI | Automatic model selection when the Prompt Router is active: receives up to 4,000 characters from the beginning and end of the latest user message, organization routing instructions/preferences, model catalog facts, and structural request parameters; returns typed model choices. Separately, per-request business-function classification (optional, separate administrator opt-in per organization) | Not published pending verification | Up to 4,000 characters from the beginning and end of the latest user message, organization routing instructions/preferences, model catalog facts, and structural request parameters; for classification: a bounded excerpt of each classified request (latest user message at most 2,000 characters, head of system prompt at most 600 characters, declared tool names), plus the API-key and team names |
| Mistral AI (La Plateforme) | AI | Model inference (first-party Mistral models — mistral-medium-3.5, codestral-2508, ministral-14b — direct La Plateforme API) | European Union — France-headquartered provider La Plateforme runs on EU-based infrastructure by def | Prompts and completions for hosted inference |
| OpenRouter, Inc. | AI | Routing layer for certain open-weight model routes (Meta Llama, Mistral). OpenRouter itself does not store request content by default; the downstream serving endpoint's policy applies | United States (routing) downstream endpoint varies | Request content routed to downstream serving endpoints |
| fal — Features & Labels, Inc. (fal.ai) | AI | Hosted video and image generation |
United States
|
Prompts and media inputs for hosted video and image generation |
| Groq, Inc. ✓ | AI | Hosted inference of open-weight models on Groq's LPU infrastructure (gpt-oss-120b-groq, gpt-oss-20b-groq, qwen-3.8-27b-groq; api.groq.com). Receives prompts and returns completions in transit | Not fixed between the United States Canada — Groq Inc. is a United States operator (Mountain Vi | Prompts and completions in transit |
| Cerebras Systems, Inc. ✓ | AI | Hosted inference of open-weight models on Cerebras wafer-scale systems (gpt-oss-120b-cerebras, qwen-3.8-27b-cerebras; api.cerebras.ai). Receives prompts and returns completions in transit | Not fixed between the United States Canada — Cerebras Systems Inc. is a United States operator | Prompts and completions in transit |
| GMI Cloud | AI | Hosted inference of open-weight serving legs (DeepSeek V4 family, GLM 5.2/5.3/5.3 Flash, Kimi K2.6/K2.7 Code — api.gmi-serving.com) | United States (San Jose CA company) serving region for our account pending verification | Prompts and completions for hosted inference |
| Prodia (Prodia Labs Inc.) | AI | Text-to-image inference (flux-2-dev, flux-2-klein-4b, flux-2-klein-9b, flux-1.1-pro, sdxl — inference.prodia.com; the FLUX1.1 pro tiers are proxied by Prodia to Black Forest Labs). Receives prompts and returns image bytes, which AI Reserve re-hosts | United States (Atlanta GA company distributed GPU network no published per-request processing reg | Prompts and returned image bytes |
| Tensormesh (serverless inference) | AI | Model inference (open-weight serving legs, serverless.tensormesh.ai) | United States pending verification | Prompts and completions for hosted inference |
| ByteDance (BytePlus ModelArk) | AI | Hosted image (Seedream) and video (Seedance) generation via the international ModelArk API | Singapore (BytePlus international ap-southeast) operator headquartered in the PRC | Generation prompts and media inputs for hosted image and video generation |
| StreamLake (Kuaishou "Vanchin" international platform) | AI | Model inference (open-weight -streamlake serving legs, vanchin.streamlake.ai) | Singapore (stated data-storage location) operator's parent headquartered in the PRC inference host | Inputs and outputs for model inference |
| Hugging Face, Inc. (Inference Endpoints) | AI | Dedicated GPU inference endpoints that enterprise administrators deploy one-click from the AI Reserve console (open-weight models, e.g. Meta Llama, Mistral, Qwen, Phi, Gemma, BGE embeddings). Endpoints are provisioned under AI Reserve's Hugging Face organization as token-protected, single-tenant deployments | United States (New York HQ endpoints in the current catalog provision on AWS us-east-1) | Prompts and completions sent to the endpoints; they receive traffic only for the organization that deployed them |
| Higgsfield Inc. | AI | Wan 3.0 text-to-video, image-to-video, and reference-to-video (wan3.0-video-higgsfield, wan3.0-i2v-higgsfield, wan3.0-reference-higgsfield). Receives generation prompts and caller-supplied image, video, and audio references by public URL and hosts generated videos | United States (San Francisco headquarters) serving location downstream AI compute providers for | Generation prompts and caller-supplied image, video, and audio references by public URL; hosts generated videos |
| Hangzhou DeepSeek Artificial Intelligence Co., Ltd. | AI | Model inference, direct DeepSeek API (deepseek-chat, deepseek-reasoner) | China (PRC) — data collected processed stored on servers in the PRC | Model inference inputs and outputs |
| Moonshot AI (Kimi) | AI | Model inference, Moonshot's own API (kimi-k3, moonshot-v1-*) | China (PRC-headquartered). International platform terms state Singapore-located servers endpoint se | Model inference inputs and outputs |
| NVIDIA Corporation (hosted NIM API catalog) | AI | Model inference (Nemotron model family) — only via the separate, clearly-labeled NVIDIA-served options (nemotron-3-ultra-nvidia; two further options, nemotron-3-nano-nvidia (retired by NVIDIA effective August 28, 2026) and nemotron-3-super-nvidia, have also been retired by NVIDIA) |
United States
|
Model inference inputs and outputs |
| Harmonic | Data Enrichment | Company-data enrichment provider that receives only public company identifiers (such as company names and website domains) — never client request content, prompts, completions, or personal data of client users | – | Only public company identifiers (such as company names and website domains) |
No subprocessors match your filters.
Your data may flow beyond direct subprocessors. Click to expand and see the chain.
Loading data flow...
No changes recorded yet. Changes will appear here after the next scan detects updates.