BLACKBOX PRESS KIT

Company news, ready to use.

The latest from Blackbox, plus company facts, leadership, product context, and media assets.

01 · OVERVIEW

Blackbox at a glance

Core company context for reporters, analysts, and partners.

Blackbox is the high-trust inference layer for teams that deploy AI at scale.

Blackbox (Blackbox AI Technologies Inc.) is the high-trust inference layer for teams that deploy AI at scale. The platform routes requests across 300+ open and closed models through one OpenAI-compatible endpoint, with third-party-verified output speed, end-to-end encryption, and zero data retention. Blackbox sells to enterprises: it runs the open-weight model that a customer chooses as a dedicated deployment, isolated to that customer. Blackbox has offices at 535 Mission Street, San Francisco.

02 · PLATFORM

One platform. Four core products.

Enterprise inference, model routing, and programmable infrastructure for development teams.

THE REQUEST PATH

01Customer applicationAny OpenAI SDK or HTTP client
02Blackbox endpointOne base URL · one API key
03Gateway layerRouting · security · metering
04Selected model or dedicated deployment300+ models · single-tenant options

03 · ENTERPRISE

Built for the security review

Privacy, control, and isolation for teams deploying AI in production.
01

Encryption

End-to-end encryption, with data encrypted at rest with AES-256.

02

Zero retention

Blackbox retains zero data and suppresses retention on routed traffic.

03

Customer isolation

Dedicated deployments run as a single tenant, isolated from every other customer.

04

PII minimization

On Enterprise, PII is removed before a prompt reaches a closed model.

05

Identity and access

SAML SSO and advanced security controls for enterprise teams.

06

Governance

Security controls, dedicated support, and deployment governance are built into Enterprise.

COMMON WORKLOADS

Multi-model inferenceDedicated open-weight deploymentsProvider failoverUsage and spend governanceSecure development workflowsHigh-throughput workloads
View privacy architecture

05 · LEADERSHIP

Meet the founders

Available for interviews on enterprise AI, infrastructure, security, and operations.
01

Company strategy, enterprise AI adoption, pricing, and the token economy.

Robert Rizk

Co-founder & CEO
Request an interview
02

Inference infrastructure, model routing, encryption, and privacy architecture.

Richard Rizk

Co-founder & CTO
Request an interview
03

Go-to-market, operations, dedicated deployments, and enterprise support.

Roger Rizk

Co-founder & COO
Request an interview

06 · BRAND ASSETS

Assets and answers

Download-ready files and quick answers for covering Blackbox.

MEDIA FAQ

What is Blackbox?+

Blackbox is the high-trust inference layer for teams that deploy AI at scale: 300+ models behind one OpenAI-compatible endpoint, plus dedicated enterprise deployments.

How is the Blackbox platform organized?+

Enterprise Inference and the Router form the production inference layer. The Agents API and CLI expose Blackbox's coding-agent infrastructure to development teams.

How does model routing work?+

Requests hit one endpoint with one API key. The gateway sends each request to the selected model, falls back when a provider has an outage, and load-balances for latency.

What is a dedicated deployment?+

Blackbox runs the open-weight model a customer chooses on dedicated infrastructure: one tenant, isolated from every other customer.

What customer data is retained?+

Blackbox provides end-to-end encryption and retains zero data. On routed traffic, retention is suppressed through provider terms and per-request flags wherever the provider API supports it.

PRESS CONTACT

Need a quote, interview, or product visual?

Press inquiries, embargoed material, fact checks, founder interviews, and asset requests route to the same team.

ENTERPRISE@BLACKBOX.AI