FAQ · SpeiroAI

An agent plans and acts across your tools — with human approval where needed. A chatbot only responds in a window.
No. Agents have defined workflow boundaries — they read and write only what each workflow needs in your systems. Sensitive outbound actions require human approval. The Speiro Agent Platform enforces these boundaries, and Speiro AI Ops continuously monitors agent behavior for anomalies.
Workflow boundaries limit what each agent can access. Human approval gates block sensitive actions until you sign off. Least-privilege scopes on all integrations mean agents can only touch what they need. A full audit trail records every agent decision and action.
We build production agentic workflows integrated into your business tools — with governance, approvals, monitoring, and audit. Claude and ChatGPT are general-purpose chat interfaces. SpeiroAI is the difference between a tool you learn alone and a solution that works for you.
No. Self-hosted runs on your cloud (AWS, GCP, Azure, or compatible). Speiro-hosted is managed by us. Either way, agents connect to your existing tools — Slack, email, CRM, accounting — via API and webhook.
You are in complete control of which actions require human approval before execution. You define the boundaries during setup.
It depends on scope, but most custom AI agents are up and running within a few weeks. A full agent platform takes longer — we'll give you a specific estimate during your discovery call based on the number of agents involved and your existing infrastructure.
Often, yes. Let us know your deadline during the discovery call — we can prioritize the highest-impact pieces first and phase in the rest.

Handing your business to an AI partner means handing over something precious: your customer information, your numbers, your way of doing things. We built SpeiroAI so you never have to choose between getting help and keeping control.
We will never sell, rent, or share your business or customer information with anyone, for any reason.
What your agents learn about your business stays inside your business. We configure commercial AI providers (such as OpenAI and Anthropic) via their API terms so your content is not used to train their public models. Your customer lists, pricing, and know-how do not become fuel for someone else's tool.
Information is encrypted in transit with TLS 1.2+ (HTTPS everywhere on our platform and integrations). Data at rest is encrypted using industry-standard AES-256. For self-hosted deployments, encryption keys are held by your cloud provider's KMS — you hold the keys. For Speiro-hosted deployments, encryption keys are managed by Speiro's cloud provider.
A small, vetted Speiro team works on your systems only when needed. Access is role-based and follows least-privilege — limited to what each task requires and recorded. No one browses your data without a reason tied to delivery or support. Within your own organization, your existing access controls apply.
Agents connect to your tools via OAuth or scoped API keys with least-privilege permissions. For self-hosted deployments, credentials live in your vault, environment variables, or KMS — wired at deploy time. For Speiro-hosted, credentials are encrypted at rest and rotated on request. Secrets are never committed to git or embedded in prompts.
When frontier models are used (Claude, GPT-family, or others), LLM inference payloads are sent to the model provider via their API. All providers are configured under their API data-processing terms — your data is not used to train their public models. For fully self-hosted deployments with local models, no data leaves your environment.
Yes. We connect securely to the systems you already trust — Slack, Microsoft Teams, Google Workspace, HubSpot, Salesforce, Zapier, n8n, and others — rather than asking you to move your business onto something new. Your operational data stays where it belongs.
It's your information. You can ask what we hold, how it's used, and request export or deletion. No hoops, no runaround — email hello@speiroai.com.
We operate in line with applicable U.S. privacy laws and handle customer data consistent with our Privacy Policy. If you operate in a regulated industry (healthcare, payments, etc.), tell us during discovery — we design deployments to meet your requirements rather than claiming certifications we do not hold.
No hostage-taking. We return what's yours and remove what we no longer need to keep. Walking away should be as clean as walking in.

Common questions

Still have questions?