You tried it and dropped it
The team played with the chat, decided "it makes things up", and that was that. The problem was task framing and missing context, not the model.
Claude Partner Network member · services track
We implement Claude inside your business
We find the processes actually worth automating, connect Claude to your systems — CRM, documents, knowledge base, code — and train your team to work with them. Not a prompting course, a working loop.
The audit takes 10 minutes · report on screen and by email · no calls, no commitment
Sound familiar?
Most companies get stuck not on the model, but on everything around it.
No idea where to start
Many processes, limited budget. What you need is not a list of 50 ideas but 2–3 scenarios with the shortest path to a result.
Our data cannot leave the building
Customer personal data, trade secrets, security requirements. This is an architecture question: what crosses the perimeter and what never does.
Engineers still code the old way
Licences are paid for, velocity is unchanged. Claude Code pays off only with a properly configured repository, rules and review process.
AI sits apart from your systems
Until the model can see your CRM, inventory and documents, it is a toy. MCP and the API make it part of the actual workflow.
Nobody owns this internally
There is no in-house AI lead and hiring one is premature. We cover that role through the adoption phase and hand it over to your team.
Services
Four tracks — from strategy to working code
We do not resell licences and we do not teach prompt writing. We build Claude into your real systems and processes.
Strategy and process audit
We work through how the company operates today and find where Claude returns the most per unit invested. You get a prioritised list of scenarios, not a list of generic ideas.
- AI Opportunity Report from a questionnaire — free
- In-depth audit of a department or the whole company
- Prioritisation by impact vs. effort
- Licence, seat and token cost modelling
- Internal AI usage policy and data-safety rules
Implementation and integration
We connect Claude to the systems your business already runs on. The model gets access to exactly what the task requires and nothing beyond it.
- Claude Team / Enterprise rollout across the company
- Corporate knowledge base and search over internal documents
- Document automation: invoices, contracts, delivery notes
- MCP servers for internal databases, services and file storage
- Integration with CRM, ERP, Google Workspace, Microsoft 365
Building on Claude
When nothing off the shelf fits, we build it. QRoot.app has shipped production code since 2011, so we support what we build rather than handing over a slide deck.
- AI agents for a specific business process
- Internal AI tools for individual departments
- Claude Code in the engineering team: rules, subagents, review
- CI and code-review integration
- Metrics: what actually got faster, and by how much
Training and ongoing support
Adoption ends not at launch but when people use it every day. That is what training and first-months support are for.
- Role-based workshops: sales, support, marketing, legal, engineering
- Prompt library for your everyday tasks
- Internal guidelines for safe use
- Coaching an internal AI lead inside your team
- Monthly support: reviews, new scenarios, cost optimisation
How it works
Five steps, and you decide at each one whether to continue
No step can be skipped, but every step can be stopped. We do not ask you to sign an annual contract before you have seen a result.
01
Free AI audit
freeA 12-question form about your company. Claude produces a report: where your automation scenarios are, what they need, and where to start.
02
Consultation, 60 minutes
UAH 4,990We work through the report against your real processes, documents and code. You get a written summary with priorities within 2 business days.
03
In-depth process audit
from UAH 19,900For when there are many scenarios and you need a map. Interviews with departments, as-is process description, solution architecture, effort and budget estimate.
04
Implementation or pilot
from UAH 49,000We take one scenario to production: integrations, access control, testing, training people. We measure the result — and only then scale.
05
Ongoing support
from UAH 15,000/moRegular sessions, scenario reviews, new tasks, token spend control. For companies already living with Claude day to day.
Pricing
Open ranges, no "contact us for a quote"
The exact figure comes after the audit — but you deserve to know the order of magnitude before you write to us.
| Service | Price | What you get |
|---|---|---|
| AI business audit | free | 12-question form plus a report with scenarios, integrations and a next step |
| Consultation, 1 hour | UAH 4,990 | Video call with a consultant plus a written summary with priorities |
| In-depth process audit | from UAH 19,900 | Department interviews, process map, architecture, effort estimate |
| Claude implementation | from UAH 49,000 | One scenario end to end: integrations, access, testing, training |
| Process automation project | from UAH 90,000 | An agent or internal tool for a specific process, with support |
| Monthly support | from UAH 15,000/mo | Regular sessions, scenario development, token spend control |
Prices exclude VAT. We work officially through КРУТ.АПП LLC, with an invoice and closing documents. You buy Claude licences directly from Anthropic — we do not resell them.
4,990UAH
per hour of consulting
What the consulting hour includes
- 60 minutes of video consulting
- Review of your brief before the call
- Work on concrete processes, documents and code — not abstractions
- A written summary with priorities within 2 business days
- A list of required licences and team roles
- Answers on security and data residency
What happens after you get in touch
01 · Request and brief
You leave your contacts; if you have already done the AI audit we pick up your report. We come back within one business day and confirm a slot.
02 · The 60-minute consultation
A video call with Marat. We look at real processes, documents and code, and answer questions. No generic slide decks.
03 · Written summary
Within 2 business days you receive notes: 2–3 priority scenarios, required licences and roles, risks, rough timelines and budget.
04 · Your call
We can implement it together, or you can take the plan and build it with your own team. Both outcomes are fine by us.
Who you will talk to
Marat
Marat
Lead Claude implementation consultant, QRoot.app
Leads Claude adoption projects for Ukrainian companies, from the first process audit to integrations running in production.
Specialises in connecting the model to a company’s real systems: MCP servers for internal databases and CRMs, agentic workflows, and Claude Code setups for engineering teams.
Consults in Ukrainian and English. Direct about architecture and limits: if a task is not a good fit for the model, he will say so in the first hour.
“The most expensive mistake in adoption is automating a process you should simply have deleted.”
FAQ
Frequently asked questions
What is the free AI audit, and what is the catch?
There is no catch. You answer 12 questions about your company, Claude analyses the answers and produces a report: which of your processes suit automation, what needs integrating, where to start. The report is yours — you can hand it to your own team and do the work yourselves. We expect some companies will want a substantive conversation afterwards, and that is sound economics for us.
Do you resell Claude licences?
No. You buy Claude licences and subscriptions directly from Anthropic. We act as a services partner: we help you choose the right plan, size the number of seats and set everything up afterwards. That is the standard model for the services track of the Claude Partner Network.
How are you different from general AI consultancies?
We focus on Claude specifically and on the engineering side of it: MCP servers, API integrations, AI agents, Claude Code in development teams. QRoot.app has been building products since 2011, so we do not only advise — we build and maintain these solutions ourselves.
Is it safe to expose our data to the model?
That is an architecture question and we cover it in the first consultation: which data leaves the perimeter at all, what can be anonymised, where an internal proxy and logging are needed. Anthropic does not train models on business account data by default — the exact terms depend on your plan, and we check them against your requirements.
Why are prices "from" rather than exact?
Because the cost depends on how many systems must be connected, what state they are in, and how many people need training. The ranges on this page are real lower bounds, not bait. You get the exact figure after the audit, in writing, before work starts.
We are a small company. Isn’t this expensive?
The audit is free and an hour costs less than a developer-day. Often it turns out you need one simple scenario your own team can ship. We will tell you plainly if a large project is not warranted.
Do you work on site?
On site in Kharkiv. Everywhere else in Ukraine and abroad we work remotely: Google Meet, Zoom or your corporate platform.
Can I reschedule or cancel?
Yes, free of charge up to 24 hours before the start. If the consultation does not happen through our fault, we reschedule or refund in full.
What if the hour shows Claude is not right for us?
That is a normal and useful outcome. We will say so directly and, where we can, point you to the tool or approach that fits your problem better.
Booking
Book a consultation
Fill in the form — we reply within one business day, confirm the slot and send the meeting link and invoice.