Backend and infrastructure for startups · Yerevan · UTC+4

Open to discussing projects

Backend and infrastructure for product launch and growth

Design and development of critical product modules: authentication, subscriptions, integrations, CI/CD and cloud infrastructure. Systems are built for reliable operation, recovery from failures and scaling as demand grows.

Core technologiesNode.js · TypeScript · PostgreSQL · Yandex Cloud · AWS
Since 2017backend and infrastructure
60–100krequests per minute under sustained load
MVP → productionlive products with real users
AWS certifiedDeveloper – Associate

Typical projects

From a single module to product launch

A project can start with a single module, a product launch or ongoing backend support. The expected outcome, scope and acceptance criteria are defined first.

01

Authentication & access control

Users, teams and organizations need secure sign-in and a clear permission model.

Sign-up and sign-in, sessions, invitations, account recovery, OAuth and resource-level permissions.

  • Sessions
  • Roles & access
  • Organizations
Discuss a similar project
02

Plans & subscriptions

As a product grows, plans, limits and available features can become inconsistent.

Individual and organization plans, subscription terms, plan changes, payment integrations and feature access.

  • 1 / 3 / 12 months
  • Product limits
  • Payments
Discuss a similar project
03

Email & user notifications

Registration, account recovery and other critical messages are often implemented as separate flows that become hard to maintain.

A single reusable system for account verification, password resets, team invitations and other service messages.

  • Templates
  • Events
  • Postbox / SES
Discuss a similar project
04

Product MVP & launch

At an early stage, architecture, data and the first releases need to evolve together.

A backend for the core user journeys, ready for launch, support and continued development.

  • Architecture
  • API + data
  • Launch
Discuss a similar project
05

CI/CD, monitoring & recovery

Manual releases and limited visibility into system health make every change risky.

Automated releases, safe migrations, centralized logs, metrics, alerts and a documented recovery process.

  • GitHub Actions
  • Docker
  • Grafana / Loki
Discuss a similar project
06

Reliable cloud infrastructure

A startup needs infrastructure that can handle failures and growth without premature complexity.

Multiple availability zones, autoscaling, managed databases, secrets, backups and background jobs.

  • Yandex Cloud
  • AWS
  • 3 availability zones
Discuss a similar project

Delivery approach

From a product goal to a stable system

Work starts with the expected outcome and constraints. Architecture decisions are explained in plain language, and the team receives the documentation and tools needed to operate the system.

01

Understand

Goal, current constraints and risks

02

Design

Architecture, data and integrations

03

Build

Development, migrations and release

04

Operate

Monitoring, recovery and knowledge transfer

Technology
Backend & dataNode.js · TypeScript · NestJS · PostgreSQL
Messaging & data exchangeRedis · Kafka · RabbitMQ · gRPC · Socket.IO
Infrastructure & monitoringDocker · GitHub Actions · Nginx · Prometheus · Grafana · Loki

Cloud choice

Yandex Cloud or AWS — based on your product needs

Architecture is designed for the selected cloud based on load, availability, security, cost and recovery requirements.

3 availability zonesAutoscalingManaged databasesBackups & recovery
Yandex Cloud
AWS
Compute CloudInstance Groups3 zones
01Compute & scaling
EC2Auto ScalingELB3 AZs
Managed PostgreSQL
DB02Relational databases
RDS PostgreSQLMulti-AZ
Managed YDB
03Distributed data
Aurora DSQLDynamoDB
Object StoragePostbox
04Files & email
S3SES
IAMLockbox
ID05Identity & secrets
IAMSecrets ManagerKMS
Cloud FunctionsCloud Backup
06Automation & recovery
LambdaAWS Backup

Selected work

Complex products running in real-world operations

Three examples of engineering challenges and outcomes. Full case studies and professional background are on the Work page.

View all work
01Aviation · operations

RUSCAMP at RusJet

The platform is in full operational use at RusJet, supporting flight managers, warehouse specialists and engineers.

Maintenance · inventory · utilization
02High load · confidential

Scenario engine

A visual scenario engine running under a sustained load of 60–100k requests per minute.

Visual rules · sustained load
03Manufacturing · full cycle

Kitchen order system

The system connected 3D kitchen configuration, designer approval, production drawings, factory progress and delivery.

3D design · factory · delivery

Discuss a project

What needs to be built in your product?

Briefly describe the current situation and desired outcome. You’ll receive clarifying questions, a realistic next step and a recommendation on hourly or project-based work.

Preferred contact method