Open to opportunities

Evgeny Nadtocheev

Fullstack Developer

Backend & DevOps & AI

I build high-load systems on Node.js and Go. I implement microservice architectures, set up infrastructure, and automate processes with AI.

5
years of experience
200K+
daily visits
2000
rps in production
10+
microservices

About Me

I am a fullstack developer and mentor with experience in small and large teams, from projects with established processes to those where processes need to be built from scratch. I helped build two startups from the ground up. My expertise spans AI development, fullstack engineering, DevOps, Kubernetes and microservice systems. I bring a product mindset and an understanding of business, connecting technical decisions to user needs, product goals and long-term development.

Technical Expertise

I build backends with Node.js and TypeScript, interfaces with React and Vue, and design microservices and integrations. I work with high-load systems, Kubernetes, CI/CD and observability through OpenTelemetry. In AI development, I build RAG pipelines and AI chats with streaming responses, integrate LLMs into products and use AI tools daily. My approach centres on clear architecture, strict typing and integration tests.

Product and Business

I have worked in fintech, real estate analytics, system integration and IT consulting. I consider the whole problem, from user needs and business logic to architecture, launch and production support. I develop my own products: SPIN, a new-build price tracker with Telegram alerts, and SMETAS, a SaaS platform for construction estimates. This experience helps me assess decisions in terms of value, cost and future development.

Teamwork and Mentoring

I have led a small engineering team, conduct technical interviews for hiring and help developers grow. As a mentor, I work through real engineering problems, architecture and career decisions. I am an expert and lecturer at School 21 (Sber) and mentor on Solvery and GetMentor. I have delivered 30+ paid and 50+ free consultations. I value clear communication, sharing knowledge and taking responsibility for the team’s results.

Working Together

I am based in Aprelevka, Moscow Region, work remotely or in a hybrid format, and am open to relocation. My English level is B2. Find my CV and detailed experience on Habr Career. The easiest way to discuss a role, project or mentoring is through Telegram.

Areas of Expertise

Key domains I work in

Backend & Microservices

Designing and building scalable high-load microservice systems

Node.jsGoNestJSFastifyExpressMoleculer.jsPostgreSQL (pgvector)MySQLMongoDBRedisRabbitMQKafkaBullMQgRPCWebSocketSSEDrizzle ORMPrismaTypeORMSequelize

Architecture

Reliability patterns and architectural approaches for distributed systems under high-load and real money

MicroservicesEvent-DrivenOutbox PatternCircuit Breaker (Opossum)SagaCQRSDDDHexagonalClean ArchitectureBFFGraphQL FederationFSDSlice ArchitectureIdempotencyOptimistic LockingGraceful Shutdown

DevOps & Infrastructure

Deploying, monitoring, and maintaining production infrastructure

KubernetesHelmFluxCD (GitOps)DockerDocker ComposeDocker SwarmDokployGitHub ActionsGitLab CITeamCitynginxOpenTelemetryTempoLokiPrometheusGrafanaSentryk6LinuxBash

Frontend

Modern SPA and SSR applications with a focus on performance and UX

ReactReact Native + ExpoVueNext.js 15 (App Router, RSC)Vike (SSR)Nuxt.jsAstroPixi.jsBabylon.jsTypeScriptTailwind (shadcn)Redux ToolkitTanstack QueryApollo ClientFramer MotionViteWebpackVitestStorybookPlaywrightCypress

Telegram & Bots

Building Telegram bots, Mini Apps, and integrations with Telegram API

TelegrafgrammYTelegram Bot APITelegram Mini AppWebhook/PollingInline Mode

AI & Automation

Integrating LLMs and AI tools into development workflows and products

LLMRAGpgvectorEmbeddingsClaude CodeCursorVector DBOpenAI APIDeepSeekPrompt Engineering

Professional Experience

Career path and key achievements

bnmap.pro

Fullstack Developer (Node.js/React)

2 years

November 2024 - present

B2B Real Estate Analytics - #1 platform in Russia · 80/20 backend

  • Maintain and evolve microservice architecture under 40K+ DAU and 2,000 rps: Event-Driven (pub/sub), Circuit Breaker (Opossum), MinIO/S3 for reports
  • Built the full observability stack: OpenTelemetry Collector (OTLP HTTP+gRPC), Tempo (TraceQL), Loki, Prometheus, 4 Grafana dashboards / 64 panels with Telegram alerts (5xx, p95-latency, RabbitMQ backlog, Redis-down, predictive disk-full); spanmetrics + servicegraph for an auto-generated Service Map
  • Migrated proxy/scraper/notification/admin/cms services from Express to Fastify - k6 latency dropped 2× at the same memory footprint
  • Built a Vike+React SSR storefront: housing catalog, premium flow with guards and paywall, price analytics charts, AI chat on an in-house RAG with streamed responses
  • Shipped Astro + Tailwind landing pages with WebGL Pixi.js animations - SSG, strong Core Web Vitals, search engine indexing
  • Set up a Fastify + Playwright + Rebrowser + Camoufox scraper for 40 sources: anti-bot bypass, fingerprinting, ghost-cursor, RabbitMQ queue, Redis checkpoint-resume
  • Introduced pgvector in PostgreSQL for embeddings and AI RAG; distributed cron on BullMQ + Redis
  • Integrated Robokassa (cards, SBP, YuMoney) with a 30-day subscription and auto-termination via a scheduler service
  • Adopted Drizzle ORM with type-safe migrations and a shared schema package; extracted shared packages - contracts (Zod DTOs), logger (Pino+OTel), telemetry (auto-instrumentation) - for the whole monorepo
  • CI/CD on self-hosted GitHub Actions runners: changed-service detection, build + push to Docker Hub, deploy via Dokploy + Docker Swarm, pre-push gate (lint, typecheck, unit + integration)
  • Testing pyramid: vitest unit, integration on docker-compose (Postgres + Redis + RabbitMQ + headless browser), k6 load tests with historical comparison
  • VitePress internal docs portal - sped up onboarding for analysts and engineers
NestJSMoleculer.jsFastifyExpressPostgreSQL (pgvector)MySQLDrizzle ORMSequelizeRedisRabbitMQBullMQgRPCWebSocketSSEZodPinoMinIO/S3grammY ReactReact Native + ExpoVike (SSR)Next.jsAstroPixi.jsBabylon.jsRedux ToolkitTanstack QueryFSDTailwind (shadcn)Framer MotionRechartsChart.jsViteVitestPlaywrightCypress KubernetesDocker SwarmDokployLensOpenTelemetryTempoLokiPrometheusGrafanaSentryGitHub Actions (self-hosted)k6

Forward LLC

Fullstack Developer (Node.js/React)

2 years 1 month

November 2022 - November 2024

Fintech (outsourced) - major bigtech, Russian Central Bank registered financial platform operator · 70/30 backend

  • In a team of 3, built a fintech product for a major bigtech - passed CBR review and was registered as a financial platform operator
  • Financial core (wallet) and payment gateway: optimistic balance locking via version, atomic Prisma transactions, idempotent operations under concurrent load with real money
  • Outbox pattern on top of Kafka (domain events, DLQ with retry/backoff via Redis, auto-cleanup) - guaranteed DB consistency, no double charges or lost payouts
  • Payment (billing) module: crypto, fiat, online payment terminals, international payment channels
  • Full frontend observability: Sentry Next.js (SSR + CSR error boundaries, performance tracing, session replay) with OTel headers - a UI error traces all the way to the specific gRPC call in the right microservice
  • Migrated the client from REST to GraphQL Federation (Apollo Client + GraphQL Codegen, Mercurius on Fastify) - one type-safe endpoint instead of 20 REST integrations; backend changes caught by CI at the codegen stage
  • gRPC contracts via buf + ts-proto: type-safe clients/servers, two-way error mapping, OTel metrics per RPC call
  • Coordinated graceful shutdown (SIGTERM/SIGINT/SIGQUIT/uncaught handlers, timeouts) + Kubernetes startup/liveness/readiness probes with real checks against Prisma/Kafka/Redis
  • Full backend observability: OTel SDK + Collector, Prometheus + Loki + Zipkin, Grafana; auto-instrumentation for Fastify/GraphQL/gRPC/Prisma/KafkaJS/IORedis/MongoDB; custom metrics and Kafka consumer lag
  • Infrastructure: Docker images, deploy via Helm charts to Kubernetes, GitOps with FluxCD, GitLab CI multistage pipeline, automatic isolated dev/test environments
  • Strict domain rules: kafka.send() only via the outbox table, Zod ENV validation on startup (fail-fast), schema-per-service in Prisma
  • FSD on the frontend for clean separation of business entities and a predictable codebase layout
NestJSFastifyExpressKoaPostgreSQLMySQL/MariaDBPrismaTypeORMKafkaRabbitMQRedisgRPCGraphQLMercurius FederationWebSocket (Socket.io)ZodHusky Next.js 15 (App Router, RSC)Apollo Client 3GraphQL CodegenReactVueRedux ToolkitTanstack QueryZustandFSDTailwindMantineReact AdminReact EmailReact Hook Formnext-internationalPlaywrightVitest KubernetesHelmFluxCD (GitOps)DockerGitLab CIOTel CollectorPrometheusLokiZipkinGrafanaELKSentryLighthousek6

Systems-fd

Fullstack Developer

1 year 4 months

August 2021 - November 2022

System Integration, IT consulting, fintech, marketing · 50/50 frontend/backend

  • Introduced TypeScript (replacing jsdoc) and Zod contracts - fewer type bugs, smart mocks for integration tests
  • Integrations with third-party APIs: Telegram, Google Maps, amoCRM, Stripe, Kinescope, Yandex Maps
  • Robokassa payment gateway (cards, SBP, YuMoney) with subscription, test and prod environments, webhook handling, and auto-termination via a dedicated service
  • Built an error tracking, classification, and handling system using NestJS's built-in architecture
  • Full Dependency Injection on NestJS for security and method reusability
  • Internal admin panel on React Admin - fast operations on financial product application statuses
  • Email module on React Email - formal email templates 2× faster to build
  • Frontend architecture refactor: useEffect + useContext → Redux Toolkit + RTK Query - significant tech debt reduction across sprints
  • i18n on next-international with Server Components and type-safe translation keys, multi-tenant for 10 currencies
  • Infrastructure: docker-compose, Portainer, Docker Swarm; VDS, GitHub Actions CI/CD; JWT + session cookies; auto-generated OpenAPI/Swagger
  • Team: 3 devs + tech lead + PM + designer; dailies, retro, stand-ups, customer demos
NestJSMongoDBMongooseRabbitMQRedisJWTOAuthPassportnginxZod ReactReduxReact QueryMaterialUIWebpackGraphQLREST DockerDocker ComposePortainerDocker SwarmYandex CloudGitHub ActionsGitLab CI

Key Projects

B2B Real Estate Analytics (#1 in Russia)

Fullstack Developer (Node.js/React), 80% backend

Build and operate microservices for a high-load platform (40K+ DAU, 2,000 rps): set up the full observability stack (OTel Collector + Tempo + Loki + Prometheus + Grafana, 4 dashboards / 64 panels with Telegram alerts), migrated proxy/scraper/notification/admin/cms from Express to Fastify (latency ↓ 2×), built a Vike+React SSR storefront with role-based paywall and an in-house RAG-powered AI chat with streamed responses, rolled out pgvector for embeddings.

40K+ DAU / 200K+ daily visits
2,000 rps in prod, latency ↓ 2× after Fastify migration
Vike SSR + Astro landings + Expo app
Distributed cron on BullMQ + Circuit Breaker
NestJSMoleculer.jsFastifyPostgreSQL (pgvector)Drizzle ORMRedisRabbitMQBullMQVikeAstroPixi.jsKubernetesOpenTelemetry

Fintech core for CBR-registered platform operator

Fullstack Developer (Node.js/React), 70% backend

In a team of 3, took a product for a major bigtech through Central Bank of Russia review until it was registered as a financial platform operator (ОФП). Designed a wallet with optimistic balance locking via version and atomic Prisma transactions, rolled out an outbox pattern over Kafka (domain events + DLQ with retry/backoff via Redis), migrated the team from REST to GraphQL Federation (Apollo Client + Mercurius on Fastify): one type-safe endpoint instead of 20 REST integrations.

CBR financial platform operator registration
Idempotent operations on real money
Outbox + Kafka + DLQ (no double charges)
REST → GraphQL Federation: 20 → 1 endpoint
NestJSFastifyPostgreSQLPrismaKafkagRPCGraphQL FederationMercuriusApollo Client 3Next.js 15 (RSC)KubernetesHelmFluxCD

Platform Engineering & Observability

Senior fullstack / Platform owner

Extracted cross-service infrastructure into shared packages (contracts with Zod DTOs, logger on Pino+OTel, telemetry with OTel auto-instrumentation) reusable across the whole monorepo. Stood up a three-level testing pyramid (vitest unit across all services and shared packages, integration on docker-compose with Postgres+Redis+RabbitMQ+headless browser, k6 load tests with historical comparison between runs). Set up CI/CD on self-hosted GitHub Actions runners with changed-service detection, build and push to Docker Hub, deploy via Dokploy + Docker Swarm.

3 levels of tests: unit + integration + k6
4 Grafana dashboards / 64 panels, Telegram alerts
Self-hosted GH Actions + Dokploy + Swarm
Shared packages reused across the whole team
OpenTelemetryTempoLokiPrometheusGrafanaSentryGitHub ActionsDokployDocker Swarmk6ZodPinoVitest

Own Products

Startups I came up with, built, and keep developing

AI & Automation

AI-first approach in development and products

AI as a productivity multiplier

I actively integrate LLMs and AI tools into everyday development and products. I use AI not as a toy, but as a tool that significantly accelerates speed and quality of work.

RAG Pipelines

Retrieval-Augmented Generation for search and content generation based on proprietary data

AI Assistants in IDE

Claude Code, Cursor - daily tools for code generation, refactoring, and code review

Vector Databases & Embeddings

Semantic search, storage, and working with vector representations of data

rag-chat.ts
const context = await db
  .select({ content: docs.content })
  .from(docs)
  .orderBy(cosineDistance(
    docs.embedding,
    await embed(query),
  ))
  .limit(5);

const stream = llm.stream({
  context,
  question: query,
});

for await (const token of stream) {
  reply.sse({ data: token });
}
LLMRAGClaude CodeCursorVector DBEmbeddingsPrompt Engineering

Algorithms & Challenges

Solving problems for fun and to sharpen algorithmic thinking

I enjoy solving algorithmic problems on LeetCode and Codewars - it keeps my thinking sharp and deepens my understanding of data structures.

Mentoring & Teaching

Sharing experience and helping other developers grow

I mentor through my own system and roadmaps via personal website, Telegram, and third-party platforms. I conduct mock interviews and career consultations for fullstack developers.

30+
paid consultations
50+
free consultations
2
years of teaching
5
mentoring platforms

What I do

  • Mock interviews for fullstack developers
  • Career consulting: resume building, interview preparation
  • Mentoring based on custom roadmaps and methodology
  • 30+ paid and 50+ free consultations completed

Telegram Channels

Running channels about development and career

Freelance Experience

Former freelancer on exchanges: a24.biz (opens in new tab), vsesdal.com (opens in new tab)

Paid Services

Career consultations and mock interviews for developers

I help fullstack and backend engineers grow their careers and confidently pass technical interviews. Format - online, exact time agreed via Telegram.

2,500 ₽ / 1 hour

Resume review, career track, stack choice, market positioning, and offer/grade negotiation prep.

What's included

  • Resume & LinkedIn/hh profile review
  • Career strategy and 3–6 month roadmap
  • Offer negotiation preparation
  • Call recording and written follow-up
Book on Telegram (opens in new tab)
5,000 ₽ / 2 hours

Full technical fullstack/backend interview with debrief: live coding, system design, theory, soft skills.

What's included

  • Live coding and algorithmic round
  • System design on a real-world case
  • Theory: Node.js, databases, architecture
  • Detailed feedback on strengths and gaps
Book on Telegram (opens in new tab)

Payment after the session. Returning clients and students - flexible terms available.

FAQ

The essentials: experience, stack, services, and how to reach me

What does Evgeny Nadtocheev do?

Fullstack Developer with 5 years of commercial experience. Specializes in backend (Node.js, Go, NestJS, Fastify), DevOps (Kubernetes, Docker, OpenTelemetry), high-load microservices, and AI/LLM integrations. Currently at bnmap.pro, the #1 B2B real estate analytics platform in Russia.

What is his tech stack?

Backend: Node.js, Go, NestJS, Fastify, PostgreSQL (pgvector), Redis, Kafka, RabbitMQ, BullMQ, gRPC, GraphQL Federation, Drizzle ORM, Prisma. DevOps: Kubernetes, Helm, FluxCD, Docker, Dokploy, OpenTelemetry, Tempo, Loki, Prometheus, Grafana, Sentry. Frontend: React, React Native, Vue, Next.js 15, Vike SSR, Astro, TypeScript, Tailwind. AI: Claude Code, Cursor, in-house RAG with pgvector.

Can I hire Evgeny as a mentor?

Yes. Paid services on this site: Career Consultation - 2,500 RUB / 1 hour (resume review, career strategy, offer negotiation prep) and Mock Interview - 5,000 RUB / 2 hours (live coding, system design, theory, detailed feedback). Also available on Solvery, GetMentor, and via Telegram.

Does Evgeny conduct technical interviews on the hiring side?

Yes. He has acted as a tech lead for a small team and regularly conducts technical interviews for hiring - dozens of interviews so far. He evaluates architectural thinking, code review, and Node.js / SQL / system design fundamentals.

How to contact Evgeny?

Preferred method: Telegram @eugene_nadtocheev (t.me/eugene_nadtocheev (opens in new tab)). Also reachable via email: johnn.hotmail@mail.ru and phone: +7 (920) 080-87-00. CV on Habr Career: career.habr.com/webappmaster (opens in new tab).

What industries has he worked in?

Fintech (payment systems, crypto, fiat, billing - team-built product registered as a financial platform operator by the Central Bank of Russia); Real Estate Analytics (#1 B2B platform in Russia with 40K+ DAU, 200K+ daily visits, 2,000 rps); System Integration and IT Consulting.

Does he work with AI and LLMs?

Yes, actively. He designed and built an in-house RAG-powered AI chat with streaming responses inside the analytics platform, and rolled out pgvector in PostgreSQL for embeddings. He uses Claude Code and Cursor daily - it is a real 2× productivity multiplier.

What is his education?

Academy of the Federal Guard Service of the Russian Federation (FSO), Oryol (2019). Specialty: Information Security of Telecommunication Systems; Automated Information Processing and Control Systems. English Level B2 certificate (2023).

Has Evgeny founded any startups?

Yes, founder of two products of his own: SPIN (podbor-minuta.ru (opens in new tab)) - a Moscow new-build price tracker with Telegram alerts on discounts from major developers (PIK, Samolet, Granel, Level, Etalon); and SMETAS (alibaba.hhos.ru (opens in new tab)) - a SaaS platform for construction estimates with a shared materials database and team collaboration.

What is his work format?

Remote or hybrid from the Moscow region (Aprelevka). Open to relocation. Available for clients across Russia and CIS. Mentoring and paid services worldwide for Russian-speaking developers (English available on request, B2 level).

What paid services does Evgeny offer on this site?

Career Consultation - 2,500 RUB / 1 hour (resume and Habr Career / hh profile review, career strategy and 3-6 month roadmap, offer negotiation prep, call recording and written follow-up). Mock Interview - 5,000 RUB / 2 hours (live coding and algorithmic round, system design on a real-world case, Node.js / DB / architecture theory, detailed feedback). Payment after the session, exact time agreed via Telegram.

Where does Evgeny publish?

Writes technical and career articles on Habr (habr.com/ru/articles/913984 (opens in new tab), habr.com/ru/articles/727810 (opens in new tab)) and vc.ru. Lectures and acts as an expert at School 21 (Sber) and at IT conferences and meetups. Runs Telegram channels @captain_galera and @eugene_vibecode.