A cleaning company platform: orders, specialists and training in one system
The company was entering the Moscow market and wanted to run on its own platform from day one. In a month we launched the MVP: a website with price calculation and ordering, client, specialist and office workspaces, and the training a specialist must complete before taking orders.
- Client
- Cleanforlife
- Industry
- Cleaning servicesMoscow and the region
- Timeline
- about a month to MVPAugust to September 2026
- Scope
- Platforms, Integrations
- Technologies
- Python (FastAPI), Next.js, TypeScript, PostgreSQL, Redis, Keycloak, Docker
[02] Task
What needed to change
A platform of its own, where the client calculates and orders a cleaning, the office assigns specialists and handles payments, and only specialists who have completed training are admitted to orders.
What was there before us
- The company had well-defined working rules: prices, cancellation and rescheduling terms, support response times, training requirements for specialists
- There was no system in which those rules could run
- Off-the-shelf booking CRMs do not handle an address with entrance and floor, access to the apartment, a checklist with before and after photos, or admission to orders based on training results
How it got in the way
- Booking, specialists' schedules, training and money would have to be kept in different services and reconciled by hand, including checking whether a specialist is cleared to work
- If a cleaner lacks the entrance, floor and intercom code, they call the client on site and lose time
- Specialists from Kyrgyzstan find instructions in Russian harder to follow, including safety and chemical handling
Why nowThe company was launching in Moscow and wanted to work on its own platform from the first day, so as to avoid migrating off an off-the-shelf CRM later.
[03] Solution
What we did
-
01
A website with an order builder
The client chooses an apartment or a house and enters the address with street and house suggestions, the area, the number of rooms and bathrooms, and can exclude rooms from the cleaning. The apartment plan redraws in front of the user, and the price is shown line by line: what it consists of.
-
02
Client workspace
Orders, rescheduling and cancellation with clear terms, properties with access details, a subscription for regular cleaning, the history of work on the home and a separate consent to use before and after photos in advertising.
-
03
Mobile specialist workspace
The route for the day, the property card, a checklist with mandatory photos, check-in on arrival and earnings. It works with a poor connection too. The access code to the apartment is visible only during the visit.
-
04
Office workstation
Tasks for today, the order log, specialist matching with an explanation of why this person is recommended, the client base, complaints with deadline control, specialist payouts and the stock of supplies.
-
05
Training and admission
Courses, tests, practice checked by a supervisor, and certificates. A specialist without a valid certificate does not appear in order matching. The supervisor has a workspace with a course builder and analytics.
-
06
Two languages for specialists
The specialist workspace, the hiring section and the training materials are available in Russian and Kyrgyz, switched with one button. The client side stays in Russian.
What the solution consists of
- Platformspublic website, client, specialist, supervisor and office workspaces, a server core with business rules, single sign-on with roles
- Integrationsan in-house address suggestion service, maps on open data, groundwork for payments, an online cash register and SMS
- Public website
- Client workspace
- Mobile specialist workspace
- Supervisor workspace
- Office workstation
- Server core with business rules
- Single sign-on with role separation
- Address suggestion service
Stages and timing
-
01
Design and core
We studied the brief and the company's regulations, designed the platform and built the core: the service catalogue, price calculation, orders, cancellation and rescheduling rules.
-
02
Website and workspaces
The website with the order builder, client, specialist and office workspaces, training with admission to orders, specialist matching.
-
03
Launch
The production domain, registration of clients and specialists, bot protection, security and backup setup.
-
04
Refinement with the client
Brand style, the address split into fields with suggestions for Moscow, mandatory property parameters, Kyrgyz for specialists.
-
05
Handover to the client
The MVP is delivered. The client runs the launch of the business, the choice of payment providers and legal and operational matters, and we stay on as technical support.
Roles
- We
- Design, development, infrastructure and the MVP launch. Technical support after handover.
- Client
- Business rules and regulations, texts and photos, acceptance. After the MVP handover: running the business on the platform, legal matters and connecting providers.
[04] How it works
What the system is made of
Integrations
- Address suggestions for Moscow and the region on open map data, with the option to connect DaData (FIAS)
- Maps on open base layers, no paid APIs
- The platform is ready to connect payments, an online cash register under 54-FZ and SMS: the client chooses the providers
Technologies
- Python (FastAPI)
- Next.js
- TypeScript
- PostgreSQL
- Redis
- Keycloak
- Docker
[05] Non-standard
What had to be solved separately
-
Admission through training
A course, a test, practice supervised by a supervisor and a certificate with a limited validity period. Without it a specialist does not appear in order matching, and the dispatcher sees why.
-
An address nobody has to ask about again
Street, house, entrance, floor, apartment and a note such as "entrance from the yard" are entered in separate fields with suggestions. The platform works out the service zone itself.
-
Our own address suggestions
The suggestion service for Moscow runs inside the platform. It does not depend on the limits and prices of external APIs, and clients' addresses never leave it.
-
Language by choice
Each specialist picks a language. If a lesson is not translated yet, they see the Russian text instead of an empty screen.
[06] Screens
What it looks like
[07] Result
What we achieved
The company has a platform of its own, where ordering a cleaning, the specialists' work, training and admission to orders come together in one place and follow its own rules. Specialists get instructions and training in their native language, and a cleaner arrives with the full address and access details. The MVP is delivered and runs on the production domain. The client runs the business on the platform, and we provide technical support.
- ≈1 mo from start to MVP August to September 2026
- 4 workspaces: client, specialist, supervisor and office
- 7 roles: client, specialist, dispatcher, administrator, owner, supervisor, HR
- 2 interface languages for specialists, with translated training materials
An honest note on impactThe business impact has not been measured yet, so we quote no percentages or payback times. The figures above describe the project itself.
Delivered
- Everything described in "What we did": the website with the builder, four workspaces, training with admission, two languages for specialists
- The production domain, registration of clients and specialists, bot protection, backups
Left as a recommendation
- Connecting a payment provider, an online cash register under 54-FZ and SMS: the client chooses the providers