FAQ
The questions that decide whether a workflow gets automated.
Short, direct answers about scope, cost, timelines, integrations, data handling and the boundaries of what I claim. If the answer you need is not here, ask it on the call.
The short version
Engagements are fixed-price, most projects sit between €1,500 and €8,000, and the first step is a 30-minute call followed by a written proposal. Start with one workflow, not a department. The audit is the cheapest way to find out whether a build is justified.
Section 01
Working together
What does an AI automation engineer do?
An AI automation engineer turns a repetitive workflow into a reliable system. That means mapping the process, measuring the hours, connecting the tools, adding an AI step only where it earns its place, and designing the human review path. It is process engineering with AI as one component, not a chatbot project.
How do we start working together?
Send a short description of one workflow, who does it today and how often. If it looks like a fit, we have a 30-minute call and I send a one-page proposal with scope, deliverables, price and a baseline. Nothing is billable until you approve it.
Do you sign NDAs?
Yes. I can sign your NDA before the first technical call, or provide a mutual NDA. I minimise data movement, work inside your environment when required, and do not publish client details without written permission.
Do you work with clients outside Romania?
Yes. I work remotely with teams across the EU and the US, from Cluj-Napoca, Romania (EU time zone). Calls are in English, Romanian or Italian. Invoicing is in EUR.
What happens after delivery?
Every engagement ends with a runbook, documentation, the measurement report and a handover session. Ongoing monitoring is available through the monthly retainer, but the default is that your team can operate the workflow without me.
Section 02
What can be automated
What can you automate with AI?
The best candidates are repetitive, high-volume steps with a clear input and output: document and invoice processing, inbox triage, data entry between systems, internal question answering, report generation and first-line support. The work is to find the exceptions before automating the happy path.
How do you decide what is worth automating?
Volume, handling time and exception rate decide it. A task that takes five minutes but happens twice a month is usually not worth an integration. A task that takes two minutes and happens 2,000 times a month is a different business case. The audit puts those numbers on one page.
Do we need to replace our current tools?
No. The workflow is built around the systems you already use, through an API, webhook or structured export. Replacing a working tool is usually a bigger project than automating the step around it.
How do you measure time saved?
On the same samples before and after: time per item, manual touches, cycle time and exception rate. The report shows a range with the assumptions written down. It does not claim a generic productivity percentage.
Is a human still in the loop?
Where the cost of a wrong action is high, or the model is not confident, yes. The review step is designed into the workflow. The goal is to remove repetitive work, not to remove judgement.
What happens when the automation fails?
Every item has a visible failure path: a review queue, a retry or an alert. Silent failure is the main risk in automation. Logs, error handling and a named owner are part of the build, not an afterthought.
Section 03
Commercial
How much does an automation project cost?
A workflow audit is €1,500 to €2,500. One automation build is €3,000 to €6,000, document and inbox automation is €2,500 to €5,000, an internal assistant is €4,000 to €8,000, and an integration path is €3,000 to €7,000. The retainer is €750 to €1,500 per month.
How long does an automation project take?
The audit is five working days. A workflow build takes two to four weeks, document automation two to three, an internal assistant three to four. The timeline assumes access to the team, the systems and a sample of real data.
Do you charge hourly or fixed price?
Fixed price for defined work. You get a written scope and a number you can approve before anything starts. Hourly is available for small change requests after handover.
Why not just use a no-code automation tool?
Use one when it fits. If the workflow needs validation, permissions, retries, an audit trail or a model with a defined output schema, custom code is usually more reliable and easier to debug. The audit can recommend either.
Section 04
Credentials and boundaries
Are you NVIDIA certified?
Yes. I hold the NVIDIA Certified Professional: Gen AI LLMs (February 2026 to February 2028) and the NVIDIA Certified Associate: Gen AI LLMs (November 2025 to November 2027). The engineering depth matters when the automation needs a private model, fine-tuning or a deployment on your own hardware.
Is our data used to train a model?
No, not by default. The work is retrieval, inference and integration. If a fine-tune is ever justified, it is a separate, explicit scope with a data agreement, not a hidden default.
Do you build chatbots?
Only as part of a knowledge or support workflow where the value is measurable. A generic FAQ chatbot is usually a worse investment than fixing the knowledge base, routing or escalation path it sits on.
Do you build full web applications?
No. I build the automation, integration and AI layer around your existing systems. If the project needs a full product build, I will say so and point you to a team that does it.
Do you provide EU AI Act compliance?
No. Legal classification and compliance opinions are a law firm’s product. I can build the technical evidence a lawyer needs: system behaviour, logging, model boundaries, test results and human-oversight hooks.
Next step
Still have a question?
Send it in an email. I answer workflow and commercial questions directly, including the ones where the answer is “leave it manual”.
Prefer email? boroicamarius01@gmail.com