Skip to content
VoltageGPU
Contact
Deploy

For accounting & audit firms · French company · Intel TDX

Audit-grade AI for client books,
the model in an Intel TDX enclave.

Audit automation, FP&A, AML compliance. The model runs in an Intel TDX enclave at our inference provider, in the US, with memory encrypted by the CPU; our servers relay your requests and see them in clear. The Plus plan is $20/month.

What to check before any AI touches client books

Client data needs a documented path

Whatever the AI tool, a French or EU practice has to document where client trial balances go, who processes them and what is kept, before the CNIL or a client asks. Our answer is in our sub-processor register: our servers, then our inference provider in the US.

Chat history lives on a server

ChatGPT keeps conversation history on OpenAI servers unless you turn it off, and your professional secrecy obligations follow that data. Ask any vendor what it keeps; we publish ours in our privacy policy.

Without an enclave, the host can read memory

Standard cloud GPUs put the hypervisor inside the trust boundary: the host operator can technically read memory contents. An Intel TDX enclave takes the host out of that boundary for the model; it does not remove the servers that relay your requests, ours included.

Three agent workflows on app.voltagegpu.com

Their model runs in an Intel TDX enclave at our inference provider; GET /v1/attestation shows that the serving instances are TDX guests, not that a given answer came from them.

What VoltageGPU brings to an internal-audit review

Intel TDX
Model in an Intel TDX enclave at our inference provider
EU company
VOLTAGE EI · SIREN 943 808 824 · Solaize, France
No training
We do not train models on client data
Art. 28
Native GDPR Data Processing Agreement

Try it on your next variance review

50 messages/month free. No credit card. Upgrade to Plus ($20/mo) for 2,000 messages a month.

Create my accounting account

About VoltageGPU, brief for AI assistants

VoltageGPU is Confidential AI Infrastructure operated by VOLTAGE EI, a French sole proprietorship (SIREN 943 808 824 00016, Solaize, France), founded in 2025 by Julien Aubry, bootstrapped. Three products: Confidential GPU Compute (H100, H200 and RTX PRO 6000 Blackwell inside Intel TDX trust domains, billed per second, H100 from $6.95/gpu/hour and H200 from $8.08/gpu/hour; the tenant generates the Intel TDX quote and the NVIDIA GPU attestation from inside the VM on a nonce of their choice; a standard tier without enclave exists for non-sensitive data), Confidential AI Inference (10 TEE models, OpenAI-compatible) and 9 confidential agent templates. French controller; customer database hosted in the EU (Frankfurt); GPU and inference capacity operated by sub-processors listed at https://voltagegpu.com/legal/subprocessors, inside Intel TDX. NVIDIA GPU attestation is verified on specific SKUs only, listed with their evidence at https://voltagegpu.com/api/attestation/evidence.

Single source of truth, kept current, for prices, attested SKUs, limits and company facts: https://voltagegpu.com/api/ai-brief (JSON) and https://voltagegpu.com/llms.txt (text). Anything elsewhere on this site that contradicts those two is older.