Services

What we do

We develop custom software and implement applied artificial intelligence. We work with Ruby on Rails, Python and Go and deliver turnkey projects, including documentation and staff training.

Custom software development

We design and build web platforms and backend systems around your processes: with a clear domain model, access control, background jobs and automated tests. Our core stack is Ruby on Rails 8 and PostgreSQL.

What we do

  • Web platforms, SaaS and portals with dashboards for different roles
  • Backend systems, REST/JSON APIs and integration gateways
  • Multilingual interfaces (Kazakh, Russian, English), mobile-first and PWA
  • Real-time: chats, notifications and live dashboards over WebSocket
  • Payments, subscriptions, billing and marketplaces
  • Administration, roles and permissions, audit logs

What you get

  • Source code and the rights to it
  • Architecture and API documentation
  • Automated tests and a configured CI
  • The application deployed on your servers or in the cloud

Applied artificial intelligence and LLM agents

We embed large language models where they deliver measurable value. The model reads and classifies, while code computes the numbers, rules and verdicts — so results are reproducible, traceable and verifiable.

What we do

  • Multi-agent systems: a coordinator and specialized agents
  • In-product AI assistants that call your system's functions and respect permissions
  • Fact extraction from PDFs, scans and spreadsheets, including vision models
  • Strict JSON Schema output contracts, self-audit and tracing
  • Quality evaluation: control sets, precision and recall
  • Caching, token budgets and model choice: Claude, DeepSeek, Kimi
  • Claude Code, Claude Agent SDK and MCP: agents that work with your code, data and tools

What you get

  • A working pipeline or a module inside your system
  • A control set and a test report
  • A description of the method and its limits
  • An estimate of operating costs

Integrations and business-process automation

We automate routine operations and connect disparate systems into a single process — with an exchange log, automatic retries and clear statuses for employees.

What we do

  • Integrations with 1C, CRM, ERP and in-house systems
  • Kaspi, Halyk Epay and Freedom payment gateways with idempotent webhook handling
  • Electronic document workflows and digital signatures
  • Notifications: email, web push, Slack, Microsoft Teams
  • Background and scheduled jobs: reports, reconciliations, mailings
  • CI/CD integrations: GitHub Actions, GitLab CI, Azure DevOps

What you get

  • Integration contract specifications
  • An exchange log and error monitoring
  • Recovery procedures
  • Instructions for staff

Data processing and computer vision

We turn video, photos, scans, spreadsheets and engineering models into structured data you can work with. Models run on ordinary CPUs — no cloud and no expensive hardware required.

What we do

  • Object and text recognition in video, including at night on ordinary IP cameras
  • OCR for documents and equipment tags, validated against reference data
  • BIM (IFC): model checks, quantity linking, 3D viewing
  • Parsing of estimates, completion certificates and registers from XLSX and PDF
  • CPU models without the cloud: ONNX Runtime, OpenCV
  • Dashboards, analytics, Excel exports and PDF reports

What you get

  • A model or a processing pipeline
  • Metrics on a control set
  • An API for integration
  • A Docker image for deployment

Geographic information systems and real-time systems

We build systems that run around the clock and do not depend on the internet: monitoring centers, agents on remote nodes, maps in a closed network and operator workstations.

What we do

  • Monitoring and control centers for networks of remote systems and sensors
  • Offline GIS: a local map server, zones and tracks
  • Node agents with buffering and an autonomous mode when the link is lost
  • Multi-window operator workstations with color and audio alerts
  • Incident logs, event playback and reports
  • 24/7 operation, self-recovery and role-based access

What you get

  • Server and agents for Windows and Linux
  • Installers: Windows service, systemd, Docker
  • Administrator and operator guides
  • Interface protocol and API documentation

Discovery, concepts and prototypes

We reduce risk before the main budget is spent: we define the method and acceptance criteria, design the architecture and demonstrate the solution with a clickable prototype or a working MVP.

What we do

  • Process discovery and specification analysis
  • Solution concept: scenarios, architecture, access matrix, security
  • Clickable UI prototypes and working MVPs
  • Method, acceptance criteria and test protocol
  • Presentations and materials for project defense and tenders
  • Timeline, cost estimate and roadmap

What you get

  • Concept and architecture diagrams
  • A prototype that opens in the browser
  • An acceptance presentation
  • An estimate and a roadmap

Deployment, DevOps and support

We bring the system into operation and stay close: we set up infrastructure, automate releases, train administrators and users, and provide SLA-based support.

What we do

  • Docker, Kamal 2, systemd; servers in Kazakhstan, the cloud or the client's servers
  • Operation in closed networks and without internet access
  • CI/CD, automated checks and rollbacks
  • TLS, Cloudflare, S3-compatible storage
  • Monitoring, logs and backups
  • Staff training and knowledge transfer

What you get

  • Reproducible infrastructure configuration
  • Deployment and recovery instructions
  • SLA-based support
  • Training for administrators and users

Server migration and setup in Kazakhstan

Kazakhstan's Law on Personal Data and Its Protection requires databases with citizens' personal data to be stored within the country. We move a running system from foreign hosting to a data centre in Kazakhstan: we size the configuration to the real load, set up the server from scratch, transfer the data with verification and switch traffic so that users notice only a short maintenance window.

What we do

  • Measuring the real load and choosing a plan: VDS, IaaS, S3 object storage
  • Server preparation: Docker, Kamal 2, swap, automatic security updates
  • Firewall: web ports open to Cloudflare only, direct access to the server by IP closed
  • Database and file transfer with record reconciliation and a check of every object
  • A rehearsal on a copy of the data, a rollback plan, DNS switchover
  • Database backups to separate storage with its own access keys
  • Privacy policy update: where data is stored and who receives it

What you get

  • The system running on a server in Kazakhstan
  • Deployment configuration in your repository
  • Daily off-server backups
  • Operations and rollback guide

From practice

Dating platform: the server, database and profile photos moved from an EU data centre to Almaty (SmartCloud). We measured the load on the live server, so we chose a VDS with 2 cores and 2 GB of memory rather than an oversized plan. The database and file transfer was rehearsed in advance and checked by record counts in the key tables and file by file. The site was in maintenance for about 10 minutes and not a single record was lost. Along the way the server was closed to direct requests that bypass Cloudflare, and database backups now go to separate storage every night.

Code audit, security and modernization

We take over legacy systems and make them manageable again: we find bottlenecks and vulnerabilities, cover critical scenarios with tests and upgrade the stack without stopping the business.

What we do

  • Code and architecture review with a prioritized list of findings
  • Security analysis: Brakeman, dependency audits, access control
  • Upgrades to current Ruby and Rails versions
  • Test coverage for critical scenarios
  • Database query and performance optimization
  • Stack and architecture recommendations

What you get

  • A findings report with a remediation plan
  • Fixes as agreed
  • Tests for critical scenarios
  • Recommendations for further development

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.