Board decks, financials and deal materials are the most sensitive files your firm produces. Here is everything your security, risk and procurement teams need to review auxi, in one place.

Get the security pack Book a demo

Using Claude or Copilot already? auxi runs alongside them, under your own agreements.

Security at a glance

SOC 2✓ Certified

GDPR✓ Compliant

External audits (ASTRA)✓ A+ scores

Deploy in your own tenant or on-premise✓ Available

Bring your own model✓ Supported

Original files changed by AINever

Compliance

Certified, audited and documented.

The standards your security review will ask about, and the paperwork behind them.

SOC 2

SOC 2 certified

auxi cloud runs on Microsoft Azure with SOC 2 certification. The report is available under NDA.

GDPR

GDPR compliant

Data residency in the region you require on dedicated deployments.

AUDITS

A+ external audit scores

Regular external audits by ASTRA, with SECTIGO protection.

CONTRACTS

DPA and retention terms

Data-handling policies, retention terms and a DPA are documented in every enterprise agreement.

Where your data goes

You choose how far inside your perimeter auxi runs.

In every option the output is the same: an editable PowerPoint file, built on your templates and checked by the same rules.

Deployment

Where auxi runs

Where the model runs

Do documents leave your perimeter?

Standard

auxi cloud on Microsoft Azure (SOC 2, GDPR)

Your chosen provider, under your agreement

Only to auxi's Azure environment and your provider

Dedicated

A private backend reserved for your firm, in your region

Your chosen provider or your own hosted model

Only to your dedicated environment

In-tenant

Inside your own cloud tenant, run by your IT team

Inside your tenant: Azure OpenAI, Bedrock or self-hosted

Local

On the user's laptop, inside PowerPoint

On the same device or your own hardware

No. Nothing is hosted or sent.

AI and model governance

Your model does the thinking. auxi does the building. You stay in control.

YOUR MODEL

Bring the model you trust

Azure OpenAI, AWS Bedrock, self-hosted Llama or Mistral, or Claude, ChatGPT and Copilot under your own enterprise agreement.

NO FRONTIER LAB

Keep data away from AI labs

With a self-hosted or privately licensed model, nothing goes to Anthropic, OpenAI or anyone else.

APPROVAL

Every change waits for a yes

Darwin proposes and you approve or skip each change. It always returns a new file, so your original is never changed.

DETERMINISTIC

Rules, not prompts

Brand rules live in auxi, not in a prompt. Change your model later and every deck still comes out the same.

IDENTITY

SSO, roles and audit logs

Single sign-on and role-based access on Enterprise, with audit logs and admin portals for your IT team.

GOVERNANCE

One admin console

Set brands, approved content, Checker rules and usage limits for every team, and store assets where you say.

Documents

The security pack, and a team to walk you through it.

Tell us what your review needs. We send the documents and, if you'd like, run a working session with your IT and security teams.

Document

How to get it

Security pack: architecture and data flows

On request

SOC 2 report

On request, under NDA

Data processing agreement (DPA) and retention terms

Included in every enterprise agreement

Your security questionnaire

Send it over and we'll work through it with you

Get the security pack Security and deployment details

Already approved Claude or Copilot?

Keep them. auxi works under the agreements you already have.

auxi connects to Claude, ChatGPT and Microsoft Copilot over MCP. Your assistant keeps reasoning under your existing terms, and auxi builds the slides on your template.

auxi vs Claude and Copilot

FAQ

Questions from IT, risk and procurement.

Does any of our data go to Anthropic, OpenAI or auxi?

With a self-hosted or privately licensed model, nothing goes to a frontier lab. In-tenant and local deployments keep documents inside your environment entirely.

Can auxi work completely offline?

Yes, in a local deployment. Paired with a model on the same device or your own hardware, there is no web service in the loop, which suits air-gapped teams.

Are our original files changed?

No. Darwin always returns a new file, and every change waits for the user's approval.

How long does a security review usually take?

We agree a timeline up front. It starts with a working session on architecture, data flows and your questionnaire, and auxi engineers work alongside yours through deployment.

See your own deck, on your brand, in 30 minutes.

Book a live demo with our team, or try auxi Pro free for 14 days.

Book a demo Start free trial

Free 30-minute demo · SOC 2 · GDPR · PowerPoint 2016+ on Windows and macOS