Особисті дані приховані

OLEKSANDR ANTONIUK
Backend Engineer
Node.js / TypeScript · Backend Architecture · Agentic AI & RAG
Ukraine, Vinnytsia region · Remote
[відкрити контакти](див. вище в блоці «контактна інформація») · [відкрити контакти](див. вище в блоці «контактна інформація»)
LinkedIn: [відкрити контакти](див. вище в блоці «контактна інформація») · Telegram: @underground_alex

SUMMARY
Backend Engineer with 5+ years of commercial Node.js and TypeScript experience. Designs and builds
multi-tenant SaaS platforms and complex APIs; owns authentication, authorization, service-to-service contracts
and cross-platform billing infrastructure.
Combines backend engineering with hands-on AI work: ships LLM-powered product features and develops
codesearch-mcp, a published MCP server for RAG-based semantic code search. Runs a spec-driven agentic
workflow (OpenSpec, Claude Code, custom review and triage skills) in daily engineering practice.

SKILLS
– Backend: Node.js (NestJS, Express), TypeScript, REST, GraphQL, WebSockets (Socket.IO)
– Architecture: SaaS multi-tenancy, DDD, event-driven design
– Messaging & Async Processing: BullMQ, RabbitMQ
– Data: PostgreSQL (incl. pgvector), MongoDB, Redis
– ORM: TypeORM, Prisma, Drizzle
– AI & LLM: OpenAI API (tool calling), agentic development (OpenSpec, Claude Code, custom skills), MCP
servers, RAG, embeddings, reranking, retrieval evaluation
– Integrations & Auth: Stripe, LiqPay, App Store / Google Play IAP, LTI, Auth0, OAuth2/JWT, Pipedrive, Google
APIs (Sheets)
– Infrastructure & CI/CD: Docker, Kubernetes API, Argo CD, Traefik, GitHub Actions, Bitbucket Pipelines, AWS
(S3, SES), Hetzner, HashiCorp Vault
– Testing & Observability: Jest, Vitest, Sentry, AWS CloudWatch
– Frontend: React, Next.js

EXPERIENCE
Backend Engineer · FlexReality Sep 2024 – Sep 2026
Development of an EdTech SaaS platform and internal AI tools
– Owned end-to-end delivery of backend features across GraphQL, REST and Socket.IO APIs used by web,
Unity-based whiteboard and mobile clients: technical decomposition, implementation, cross-team
coordination and staged, feature-flagged releases.
– Redesigned platform authentication and authorization: replaced a monolithic 473-line Express middleware
with declarative, type-safe per-route policies (~316 across 172 files) and an exhaustive dispatcher that
enforces policy coverage at compile time.
– Standardized service-to-service authentication across four services in two regions: canonical header contract,
compile-time-validated key registry, timing-safe comparison and per-route caller allow-lists; cut over in a single
coordinated release, avoiding a dual-format transition window.
– Extracted cross-platform billing into a standalone service and owned the full subscription lifecycle across
native IAP (App Store, Google Play), Stripe and LiqPay, including provider webhooks, idempotency and
subscription state reconciliation.
– Shipped an in-product AI assistant with a frontend engineer: OpenAI tool calling behind built-in prompts for
lesson preparation and formula generation, and integrated a self-hosted image generation model as an
assistant tool. Evaluated OpenAI, Gemini, Mistral and self-hosted Ollama models behind a feature flag before
production rollout.
– Built and ran a spec-driven agentic development workflow on OpenSpec and Claude Code: RAG-backed
exploration, automated spec validation, multi-agent diff review, finding triage and automated fixes. Delivered
the auth redesign and the platform-wide backend i18n through it.

– Built LTI integrations with Moodle and Ukrainian school systems (NZ.UA, HUMAN), including SSO via the LTI
launch flow. Synced users, school registrations, paid-service requests and billing data with Pipedrive CRM,
replacing manual sales entry, and automated reporting into Google Sheets.
– Modernized the legacy codebase through a platform-wide logging refactor and Sentry for production error
tracking.
– Introduced the workflow and codesearch-mcp (see Projects) to the team; reviewed backend code and
onboarded new engineers.
Software Engineer · Maven Group Jan 2024 – Jun 2024
Kubernetes cluster and resource management platform
– Developed full-stack (React/Node.js) Backstage.io plugins for managing Kubernetes clusters and their
resources, deploying cluster objects programmatically via the Kubernetes API.
– Implemented REST APIs integrating Backstage plugins with Kubernetes and Argo CD, automating container
image builds from GitHub repositories and subsequent Kubernetes deployments through Argo CD.
Solution Architect / Software Engineer · Rist School Apr 2023 – Jan 2024
Multi-tenant CRM platform for education
– Sole engineer responsible for the product: architecture, backend, infrastructure and deployment of a
multi-tenant CRM built from scratch (NestJS, Next.js, PostgreSQL, AWS S3 and SES, HashiCorp Vault).
– Owned core architecture decisions (tenancy model, data isolation, auth) and covered the frontend where
delivery required it.
– Ran the end-to-end delivery cycle: business and system analysis, requirements into technical tasks, CI/CD on
Bitbucket and deployments across environments.
Node.js Developer · QATestLab Jan 2022 – Feb 2023
– Built a corporate Skype bot (Node.js, TypeScript, Microsoft Bot Framework) integrated with internal company
systems, automating internal workflows and daily team communication.
Node.js Developer · BezBrendu Jan 2021 – Oct 2021
Development of a CRM system for online sales
– Engineered a custom SDK for the Facebook and Instagram Marketing APIs handling external data
synchronization, alongside server-side logic and RESTful APIs in Node.js.
– Designed relational schemas (MySQL) and scripted Auth0 tenant synchronization for permissions and roles.

ADDITIONAL TECHNICAL EXPERIENCE
Founder & Hardware Engineer · Portal Sep 2005 – Jan 2021
– Programmed and flashed microcontrollers, EEPROM and NAND memory; reverse-engineered proprietary
hardware and embedded systems to resolve low-level failures.

PROJECTS
codesearch-mcp npmjs.com/package/codesearch-mcp · independent project, Mar 2026 – Present
– MCP server giving AI coding agents semantic search across local repositories: AST-aware chunking via
ts-morph, PostgreSQL + pgvector through Drizzle ORM, incremental re-indexing, pluggable local or cloud
embedding and reranking backends.
– Nine MCP tools: semantic search, symbol lookup, call graph and cross-project dependency graphs; Nx support.
– Retrieval evaluation harness on frozen-commit corpora (153 labeled queries; hit@k, MRR@10, per-stage
metrics, paired sign tests) backs embedding, reranking and threshold decisions; chunking changes are verified
by a separate chunk-boundary audit.
Fix Flow CRM multi-tenant SaaS for repair shops, built on hands-on repair-business experience (in development)
– Multi-tenant backend on NestJS, GraphQL and Prisma: per-tenant data isolation, role-based access (owner /
admin / finance / client) and encrypted credential storage in PostgreSQL.
– Deployed as a single-host Docker Compose stack (Hetzner, Traefik, PostgreSQL, Redis) with staging.

EDUCATION
Vinnytsia Technical College, Radio Engineering
Languages: English – Intermediate (B1) · Ukrainian – native