Skip to the content
Process

From brief to support
with no surprises in the middle.

Four stages. Each one sets out what happens, what you get in your hands, how long it takes and what we need from you. If a stage goes off plan, you read about it in the next weekly report.

I

Briefing and analysis

2–3 days

We get on a call. We go through the task, work out the economics and agree the KPIs. The first question is always the same: what has to change in the numbers. Sometimes the answer is that the task does not pay for itself, or that configuring a service you already own would do the job. Then we say so, and we do not take the project.

What you get
  • A completed brief with structured answers
  • The arithmetic: what the process costs now and what it will cost after
  • The list of KPIs we will measure the result against
  • A line-by-line estimate and a range for the timeline
What we need from you
  • One person on your side who makes the decisions
  • Access to data about the current process, even approximate data
  • An hour for a call and an hour to read the estimate
brief.json
{
  "service": "ai-agents",
  "budget": "₽300–700K",
  "timeline": "2–3 months",
  "kpi": {
    "current": "900 requests/day",
    "target": "60% without an operator"
  }
}
II

Architecture and prototype

3–5 days

We draw the diagram, build a clickable prototype and pick the stack. Before the first line of business logic you can already see how it all fits together and how it is used. Changing the structure on a prototype costs hours; changing it in finished code costs weeks. We tell you what the chosen stack cannot do straight away, rather than waiting for you to run into it.

What you get
  • A clickable prototype on a staging domain
  • An architecture diagram and the data model
  • A list of integrations with their limitations
  • A week-by-week plan of the work
What we need from you
  • Click through the prototype and mark where the path is awkward
  • Sign off the structure before development starts
  • Start collecting content and access credentials
architecture.md
client → CDN → Next.js
                  ↓
             Go API ── Redis
                  ↓
            PostgreSQL
                  ↓
            queue worker
III

Development in sprints

1–4 weeks

We write, we show, we fix. The sprints are short. Standard blocks come from our own library, and the time that frees up goes into the part that makes the product yours. Every line is read by a person before merge.

What you get
  • A demo on a staging domain every 2–3 days
  • Repository access from day one
  • Automated tests for the critical paths
  • A weekly report: done, next, risks
What we need from you
  • Look at the demos and send corrections the same day
  • Hand over access to the CRM, the payments and the domain
  • Requirements hold for the length of a sprint; if they have to change, the deadline moves with them
sprint.log
feat(api): handlers/lead.go
review: 3 comments, validation
test: lead_test.go, 42 cases
ci: tests passed 42/42
approve → merge → demo
IV

Launch and support

2–5 days + a month

We deploy, set up monitoring and hand over the credentials. Then we spend a month fixing whatever surfaces. The monitoring is there so the system tells you about a problem before your customers do. After that month the call is yours: a subscription, or your own team.

What you get
  • A production environment with monitoring and alerts
  • Source code and documentation in your repository
  • Deployment and rollback instructions
  • Training for your team
  • The acceptance certificate and the closing documents
What we need from you
  • Accept the work against the agreed KPIs
  • Name the person who will receive the alerts
  • Decide whether you want a support subscription
deploy.sh
$ docker compose up -d
 ✔ postgres   healthy
 ✔ redis      healthy
 ✔ api        healthy
 ✔ web        healthy
$ curl -s /health → 200 OK
Outcome

What you are left with
at the end.

The list is the same for every project. None of it is billed separately.

  • 01A completed brief and the arithmetic behind it
  • 02A clickable prototype and an architecture diagram
  • 03Repository access from day one
  • 04A demo on a staging domain every 2–3 days
  • 05Automated tests for the critical paths
  • 06Production with monitoring and alerts
  • 07Source code and documentation in your repository
  • 08Deployment and rollback instructions
  • 09The acceptance certificate and the closing documents
Guarantees

What we promise
and will put in the contract.

A month of free fixes

For a month after the acceptance certificate is signed we fix anything that does not match the agreed scope, free of charge. That includes the parts that work but work awkwardly.

The source code is yours

The code goes into your repository together with the exclusive rights. You will never buy a licence for your own product.

NDA on request

We sign it before the brief if the task is sensitive. Client data stays inside the agreed perimeter and never turns up in other projects.

No payment for an estimate

The brief, the analysis and the estimate are free. You start paying once the scope and the timeline are agreed.

It all starts
with the brief.

Six steps, ten minutes. After that we come back with an estimate of the timeline and the cost. It is free and commits you to nothing.

Fill in the briefSee the cases