FAQ
Straight answers before you start.
Cost, timelines, AI accuracy, data security and how we work. If your question isn't here, include it in your quote request.
01
Getting started
What does Orchento do?
Orchento is an AI business automation agency. We design, build and manage workflows that connect the tools you already use, remove repetitive work and keep your team in control of important decisions. AI is used where it helps, for example reading messages or drafting documents, while business rules, validation and approvals run in ordinary code.
Who do you work with?
B2B businesses with recurring manual processes, such as service providers, installers, consultancies, distributors and commerce teams. The quote form offers company-size bands from 1–10 people to more than 1,000 so we can scope the work to your situation.
How much does a project cost?
It depends on the process, the number of systems involved and the safeguards required, so we don't publish fixed prices. After a short conversation we send a scoped proposal, and pricing is confirmed in writing before any work starts.
How long does a project take?
Timelines depend on the integrations, the quality of your data and how quickly approvals happen. We give an estimate after scoping, and we prefer a small pilot first so you see a working result early.
What happens after I request a quote?
We review your request and reply to the business email you provided with questions or a scoped proposal. Requests are reviewed by a person; nothing is sent to your systems and no work starts without your agreement.
What should I include in a quote request?
Your business name and size, what you want to automate, how often it happens, who owns the process and which tools are involved. The more specific you are, the more accurate the proposal. Please don't include passwords or sensitive personal records.
02
Working together
Do we need to change our current software?
Usually not. We start by evaluating your existing tools and the secure integrations they support. The aim is to improve the workflow without unnecessary migrations.
Can we start with just one process?
Yes. A focused pilot is our preferred starting point: one defined workflow, clear acceptance criteria and a measurable baseline. We expand only when the first result is proven.
What do you need from our team?
A process owner who can describe how the work is done today, access to the relevant systems through accounts you control, sample data and someone to approve consequential actions during testing.
Do you provide ongoing support after launch?
Yes. Automations need monitoring: we watch executions, handle exceptions, review the audit trail with you and improve the workflow as your process changes. Support terms are agreed in the proposal.
Who owns the automation and the data?
Your data stays yours. Ownership and licensing of the delivered work are agreed in the proposal before work starts, so there are no surprises later.
Do you work with international businesses?
Yes. Orchento is an international agency and the service is delivered in English. Data protection, invoicing and system requirements are assessed per country during scoping.
03
AI and human control
What if an AI step makes a mistake?
We use validation, confidence thresholds, logged executions and human review. Sensitive changes can require explicit approval before anything is sent or updated, and uncertain results go to a review queue rather than changing your records.
Can the AI set prices or change financial data?
No. Prices, totals and other financial values are calculated by deterministic code from validated catalogs and rules you approve. AI can help prepare a draft, which a person reviews.
Will the AI send emails or messages to customers on its own?
Not unless you decide that for a specific, low-risk case. By default outbound messages are drafted and held for approval by a named person.
Is our data used to train AI models?
We review the data terms of the AI provider for each workflow, prefer configurations that do not use your data for training and document what leaves your systems before go-live. We only send the fields a step actually needs.
Which AI models do you use?
The model is chosen per workflow based on the task, data sensitivity and cost, and AI is kept to bounded jobs such as classification, extraction and drafting. We are not tied to one provider.
04
Security and data
How do you handle business data?
We assess permissions, data minimization, retention and relevant privacy obligations before integrating systems. Requirements depend on the specific process and the vendors involved.
What access do you need to our systems?
The minimum needed, through dedicated accounts or API credentials with limited permissions that you can revoke at any time. Credentials are stored server-side and never placed in prompts or in the browser.
Are you GDPR compliant or certified?
We design for lawful basis, data minimization, retention and deletion, and we agree data-processing responsibilities before touching live customer data. We do not claim any security certification we do not hold, and we will tell you plainly what applies to your project.
Can we see what the automation did?
Yes. Executions, approvals and exceptions are logged so your team can review what happened, who approved it and what needs attention.
05
Integrations and custom solutions
Which tools can you connect?
Anything with a supported API, webhook or secure data export, including CRMs, ERPs, email, calendars, spreadsheets, commerce platforms and internal tools. We confirm what is possible during discovery rather than promising an integration up front.
What if a system has no API?
We assess supported exports, secure file exchange or a controlled on-premise connector where appropriate, and tell you honestly if a reliable integration is not practical.
What is a custom MCP server?
The Model Context Protocol (MCP) is an open standard that lets AI assistants call tools and read data from other systems. A custom MCP server is a small, controlled service we build that decides exactly which of your data and actions an assistant can use, for example to produce reports or oversee operations. See our custom solutions page for details.
How is this different from a chatbot?
A chatbot talks. Our workflows act inside your business processes: they read inputs, apply your rules, update systems and ask for approval when it matters. A chat interface can be one way to work with them, but it is not the product.
Didn't find your answer?
Tell us about your process in the quote form and we'll answer in our reply.