Where a mid-size business loses money
A mid-size business loses money at the joints between departments. Sales take a request, the warehouse checks stock, accounting issues the invoice, and at every handover the work waits for a person.
Each handover costs time: a manager waits for the warehouse, a customer waits for an invoice, an executive waits for the report at the end of the week. In the loss map we count these waits in hours and money for each department.
Each department often buys AI on its own: one subscription for marketing, another for sales, a third for support. The data in them does not match, and nobody sees the combined effect.
The supplier is also chosen without comparison. According to the same survey, 91% of Russian companies work with one AI supplier and have nothing to compare its price and quality with.
Typical points of loss: processing requests and invoices, answers to customers, reconciliations, reports for the executive, preparing documents. The check shows which of them cost money in your company.
AI due diligence by department
For a mid-size business the check covers every department. AI due diligence, a full check of the company before decisions on AI, answers one question: where AI will bring money and where it will bring risk.
We interview the head of each department and review processes as they really run, not as the rules say.
We check the data and its quality, the systems (CRM, 1C accounting, phone lines), and the costs of people and contractors.
Separately, we look at what has already been done with AI and which contracts are signed with suppliers. Some of what you bought can be switched off, some can be tuned.
The check also gives the opposite result: a list of processes where AI is not needed now. For a mid-size company this protects against purchases nobody will use.
At the end you know for each department what to invest in and what not.
The map and the effect calculation
The map gathers the results of the check into one document. For each process: losses in money, solution options from the market, their price and effect.
We calculate the effect by one formula: staff on routine work, salary with taxes, the share of work for AI, minus support. The one-time cost of rollout is subtracted in the yearly calculation.
Each solution shows three numbers: what the rollout costs, how much it saves a month and in how many months it pays off. Solutions with a long payback go to the end of the plan or drop out of it.
The map has a risk section: what can be given to AI, where to store data under 152-FZ, Russia’s personal data law, and how not to depend on one supplier. For a company that has one AI supplier today, this is often the first question.
We link solutions across departments: one process from request to payment on shared data, not six different services. This is how an AI department of 3–5 AI agents works, from $4,999, 4–6 weeks.
The plan for a quarter
The map for a mid-size business ends with a plan for the quarter: what we roll out first, what second and what we postpone. Each item has a budget, an owner and a metric.
A pilot usually comes first: one AI agent on one process, from $1,499, 14 days, with the metric in the contract. When the pilot has shown a number, we connect the next processes from the map.
If a solution does not show a number, we switch it off and move to the next item on the map. This way the quarter’s budget does not go to what does not work.
The plan also covers preparing people: which employees work with the solution and what they need to know before launch.
What such a plan includes and how to run it is covered in more detail in the article on the quarterly AI rollout plan.
Who is responsible for AI in the company
Without an owner, AI belongs to every department and to none. In a mid-size company an in-house AI department solves this: the people and rules that connect departments and answer for the result in money.
The AI director answers for the map and the result and reports to the top executive. Department heads own the processes, and each has a number. The IT department connects solutions to systems and answers for access.
The lawyer and security answer for data, 152-FZ and contracts with suppliers.
The difference in terms matters: an in-house AI department is your people and rules, while an AI department from us is a team of AI agents that we build for your processes.
If it is too early to hire an AI director on staff, we take on the role: from $999 a month, 5 hours a week, with no new hires.
Support of solutions after launch: from $480 a month, cancel in any month.