About

Zhubanov SoftDev — a full-cycle engineering team

We build custom software and implement applied artificial intelligence for businesses and the public sector in Kazakhstan and abroad.

Zhubanov SoftDev (Sole Proprietor Zhubanov R. Zh.) is a software development company. We build web services and backend systems, integrations and business-process automation, solutions based on large language models and agent pipelines, data processing, computer vision, geographic information systems and real-time systems.

Our portfolio includes education and e-commerce platforms, HR systems with multi-agent AI, fintech solutions, construction and BIM tools, field reference tools for industrial engineers, security and recognition systems, and modules for accounting and government systems.

We help companies automate routine operations, speed up workflows and make data-driven decisions. We deliver turnkey projects — discovery, design, development, testing, deployment and support — including documentation and training for the client's staff.

Legal form
Sole Proprietor Zhubanov R. Zh.
Portfolio
19 projects · 14 industries
Technology
Ruby on Rails · Python · Go · TypeScript
Geography
Kazakhstan and worldwide, remote (UTC+5)
Working languages
Қазақша · Русский · English

Approach

AI you can trust

The model reads and classifies — deterministic code computes the numbers and decisions, and an expert approves the result. That is why every number is reproducible and traceable to its source.

input

Data

Documents, scans, spreadsheets, BIM models, video streams, accounting systems.

LLM

The model reads

Extracts facts, classifies, turns requirements into strict schemas.

engine

Code computes

Arithmetic, rules and verdicts live in a deterministic engine.

human

An expert approves

Ambiguous cases go to a specialist; decisions are recorded.

output

Result

A report, an API response or an action — traced back to the source document.

The model reads — code computes

We never trust arithmetic, rules or verdicts to the model. That makes results reproducible and verifiable.

Expert in the loop

Doubtful cases are never published automatically: a specialist reviews them and decisions are recorded.

Traceability

Every number leads to its source: a document, row, cell, model element or transaction.

Works without the cloud

CPU models, offline maps, closed networks — client data never leaves the client's infrastructure.

Quality by default

Automated tests, linters, security checks and CI are part of every project.

Predictable AI costs

Caching, token budgets and per-agent cost accounting — no surprises on the invoice.

Quality

Engineering standards by default

What is usually sold as an add-on is part of every project we deliver.

Automated tests

Model, service, integration and system tests — from a hundred to a thousand tests per project.

Security

Static analysis (Brakeman), dependency audits, access control, action audit logs and masking of sensitive data in logs.

CI/CD

Linters, tests and security checks on every change; reproducible deployment with Docker and Kamal.

Documentation

Architecture, admin and user guides, API docs, test reports and traceability to the requirements.

AI cost control

Request caching, token budgets, per-agent cost accounting and fallback model chains.

Confidentiality

We work under NDA, deploy on the client's servers without the cloud and never use real data without consent.

Engagement models

Choose how we work together

A contract with a registered company, a transparent estimate and clear acceptance points.

Fixed price

Turnkey project

Fixed scope, timeline and price based on an agreed specification.

  • Discovery and specification
  • Stage-by-stage acceptance
  • Warranty period
Prototype · MVP

Rapid prototype

Validate a hypothesis in a few weeks: a working prototype, metrics and a test report.

  • Method and success criteria
  • Clickable or working prototype
  • Presentation for sign-off
Retainer

Evolution and support

A monthly budget of hours for product development, support and monitoring.

  • Incident SLA
  • Updates and security patches
  • Priority access to the team

Team

Founder

Ruslan Zhubanov

Founder · Applied AI Engineer · Full-stack Ruby on Rails

More than five years in commercial software development. Designs LLM systems and takes them to production — agent pipelines and generative features inside products — along with the products themselves: the domain model, background queues, deployment and operations.

Working principle: the model reads and classifies, while deterministic code computes the numbers and verdicts — with strict call schemas, caching, self-audit and budgets.

How we work

Turnkey delivery in six clear stages

Every stage ends with a measurable result you accept before we move on.

  1. 01

    Discovery

    We study processes, documents and constraints, and define goals and success criteria.

  2. 02

    Design

    Architecture, scenarios, a UI prototype, a plan and an estimate — you see the solution before development starts.

  3. 03

    Development

    Short iterations with demos and a staging environment from the first weeks.

  4. 04

    Testing

    Automated tests, AI control sets, security checks and acceptance testing.

  5. 05

    Deployment

    On your servers, in the cloud or in a closed network — with monitoring and backups.

  6. 06

    Support

    SLA-based support, product evolution, documentation and staff training.

Have a challenge? Let's talk

Describe the process you want to automate or the product you need to build. We reply within one business day with questions and a proposed engagement format.