Senior Frontend Engineer
OleksandrKukhtin
I build interfaces that hold up under real production load — media tooling at Netflix, AI writing surfaces at Writer, and commerce systems for millions of users. When a feature needs a simple backend, I can write the Node/Express/Postgres CRUD layer for it; large-scale distributed backend design is not my specialty.
- Based
- Warsaw, Poland · Remote preferred
- GitHub
- github.com/alexisinwork
- linkedin.com/in/alexkuhtin
12+
years shipping product UI
20%
fewer incidents and bugs at Netflix
30%
manual steps removed by automation
32%
less agent time per customer at Tourlane
01 · Edge
What sets me apart from other UI engineers
Six things I bring that most frontend candidates do not — each one is demonstrable on this site, not a bullet on a slide.
- 01
Frontend ownership with basic backend support and product/UX thinking
I primarily own the UI and can write the basic Node/Express endpoints and PostgreSQL tables a feature needs. I also handle the product-management side — requirements, scope and delivery — and my earlier career as a UX designer means deep, fluent collaboration with Design and PM instead of hand-offs.
- 02
Performance treated as a product feature
This site ships its own live Web Vitals panel — LCP, CLS, INP measured in your browser, with the exact techniques used listed underneath. I hold the same budgets in production code.
- 03
AI engineering, certified and applied
AI Engineering Basics certified, with real work in RAG, embeddings, evals, tool calling and MCP — plus daily use of Claude Code, Codex, Antigravity and Kimi K3 as an engineering harness.
- 04
I build my own tools
The interview trainer and AI roadmap on this site are tools I built for myself and kept maintaining — hundreds of questions, a sandboxed code runner, and a learning graph, all self-hosted here.
- 05
Security and accessibility by default
Strict headers, sandboxed execution, server-side rate limiting, secrets never in the browser, keyboard-complete UI and honest semantics — checked, not claimed.
- 06
Netflix-scale production instincts
20% fewer incidents and 30% fewer manual steps on tooling used by thousands of productions. I optimise for the on-call engineer as much as the end user.
02 · Experience
Twelve years of shipping product interfaces
Senior Frontend Engineer with 12+ years building scalable, user-centric web applications for product teams at Netflix, Writer, HelloFresh, Zalando and Tourlane. I am primarily a frontend specialist; I can also build the basic Node.js, Express and PostgreSQL CRUD services and backend-for-frontend layers a feature needs, but I am not a distributed-systems backend engineer. I specialise in thoughtful user experience, robust frontend architecture and remote-first, transparent engineering — async workflows, sharp documentation, and no-ego code review.
Netflix
Senior UI Engineer (Frontend)Aug 2023 — Present · Warsaw, Poland (on-site)- Redesigned and developed internal drive-based workflow tools, improving file upload/download efficiency for thousands of productions, reducing incidents and bugs by 20% and manual steps by 30% through automation.
- Re-architected the backend-for-frontend layer, migrating from Node.js/Express to a Netflix-internal BFF solution and doubling system observability and monitoring capability.
- Built and maintained Node.js services and GraphQL/REST endpoints over relational (PostgreSQL) and internal data stores, owning schema changes, migrations and query performance.
- Developed a video diffing tool that visually surfaces differences between video file versions in media players and timelines, streamlining quality assurance.
- Collaborated cross-functionally with Design, Product and Support to understand user needs and ship impactful improvements.
- Championed clear asynchronous communication, documentation, and no-ego code reviews focused on clarity and shared learning.
- React
- TypeScript
- GraphQL
- Node.js
- Express
- PostgreSQL
- Jest
- Sentry
- Jenkins
- Git
- SRE
- AI-First Development
Writer
Senior Frontend EngineerDec 2021 — Jul 2022- Engineered an AI-powered writing assistant for teams: terminology management, writing rules and dynamic style-guide enforcement.
- Designed and implemented a custom Quill-based AI-assisted editor highlighting grammar, clarity and compliance issues in real time.
- Built and maintained the CMS for user and team management, supporting scalable collaboration and admin workflows.
- Extended Node.js/Express API endpoints backing the editor and admin surfaces, including data modelling in PostgreSQL.
- React
- React Hooks
- TypeScript
- Node.js
- Express
- PostgreSQL
- Lerna
- Jest
- GitHub Actions
HelloFresh
Senior Frontend EngineerDec 2020 — Jan 2022- Owned delivery of critical updates to the Delivery History feature serving millions of users.
- Consolidated three internal CMS systems, streamlining menu publishing, recipe editing and menu creation.
- Led requirements gathering, user stories, delivery coordination and documentation for multiple high-impact features as Delivery Lead.
- Interviewed frontend candidates and drove async transparency across stakeholders.
- React
- Redux
- React Query
- TypeScript
- Cypress
- Testing Library
Tourlane
Senior Frontend EngineerJun 2019 — Nov 2020- Delivered internal productivity tools that cut agent time per customer by 32% and raised customer satisfaction.
- Built MVP frontends for backend teams, accelerating new app launches and integration across Backend, Product and Design.
- Wrote Node.js/Express service and BFF endpoints with PostgreSQL-backed data to unblock frontend delivery.
- Applied TDD throughout and kept cross-functional communication open and explicit.
- React
- JavaScript
- Node.js
- Express
- PostgreSQL
- Styled Components
- Cypress
- TDD
- GitHub Actions
Zalando SE
Frontend EngineerMar 2016 — May 2019- Developed an application that streamlined campaign setup for advertising managers, increasing team capacity and contributing to revenue growth.
- Migrated the legacy platform from Angular to React, improving performance and maintainability.
- Mentored junior developers and aligned OKRs with UX, Product Management and leadership.
- JavaScript
- React
- Angular
- Redux
- CSS
- Jenkins
- Protractor
Freelance
Frontend Engineer (Remote)Feb 2015 — Feb 2016- Built client-facing web applications end to end, from requirements to delivery.
- Introduced component-driven architecture and end-to-end testing to small product teams.
- JavaScript
- React
- Angular
- Redux
- CSS
- E2E Testing
03 · Toolkit
What I work with
Core
- React
- TypeScript
- JavaScript
- React Hooks
- Redux
- React Query
- GraphQL
Craft
- Design systems
- Accessibility
- Animation / motion
- Styled Components
- Tailwind CSS
- Web performance
Quality
- Jest
- Cypress
- Testing Library
- Protractor
- TDD
- Sentry
Backend
- Node.js
- Express
- REST APIs
- GraphQL servers
- Backend-for-frontend (BFF)
- PostgreSQL
- SQL & schema design
- Migrations
- Auth & sessions (JWT/OAuth)
- Docker basics
Platform
- Electron
- Webpack
- Vite
- Jenkins
- GitHub Actions
- Git
- SRE
- Observability
AI
- LLM APIs
- AI-First Development
- Model Context Protocol (MCP)
- Tool calling
- RAG & embeddings
- Chunking strategies
- Evals
- Prompt engineering
- Claude Code
- Codex
- Antigravity
- Kimi K3
- Cursor
Leadership
- Mentoring
- Project management
- Requirements & delivery lead
- Hiring & interviewing
- Async documentation
04 · Backend
Frontend engineer with basic backend support
I am primarily a frontend engineer. When a feature needs a simple API, I can build the Node.js/Express layer behind it — CRUD endpoints, basic PostgreSQL schemas, migrations and auth — enough to ship the feature end to end. I do not design or run large-scale distributed systems.
Node.js & Express services
Basic REST and GraphQL endpoints, backend-for-frontend layers that shape data for the UI, request validation and structured logging. Not production backend architecture at scale.
- Express routing
- BFF pattern
- Zod validation
- Basic CRUD
PostgreSQL & data modelling
Relational schema design for small-to-medium features, migrations, query performance and pragmatic row-level access rules.
- Schema & indexes
- Migrations
- Query performance
- Row-level security
Production concerns
Auth and sessions, rate limiting, basic Docker-based local parity, and enough tracing to keep frontend features observable. I do not design distributed queues, caches or orchestration.
- JWT / OAuth
- Rate limiting
- Docker basics
- Tracing
05 · AI engineering
My path into AI engineering
The pipeline I follow when I take an AI feature from idea to production — from picking a model and its pricing strategy through chunking, embeddings, RAG, evals and shipping.
AI Engineering Basics
Certification · passed · 2026
Model selection, prompt and context design, embeddings, chunking, retrieval-augmented generation, evaluation and deployment trade-offs.
- 01
Frame the task
Define the job before the model: inputs, outputs, latency budget and the eval that proves it works.
- Use case & constraints
- Golden test set
- Success metric
- 02
Choose the model
Match capability to task — reasoning depth, context window, multimodality, tool calling and licensing.
- Frontier vs small model
- Context window
- Tool calling support
- 03
Pricing & hosting strategy
Decide where inference runs. Managed public APIs win on speed to market; open weights win on unit cost, privacy and control at volume.
- Public API (per-token, zero ops)
- Serverless inference provider
- Open-source weights self-hosted / local
- Cost per 1M tokens vs GPU hour
- Caching & batching to cut spend
- 04
Prepare the knowledge
Ingest, clean and chunk source documents so retrieval returns coherent, self-contained context.
- Parsing & cleanup
- Semantic vs fixed chunks
- Overlap & metadata
- 05
Embeddings & vector store
Embed chunks, index them, and pick a store that fits scale and filtering needs.
- Embedding model choice
- Vector index (HNSW / IVF)
- Hybrid keyword + vector
- 06
RAG retrieval
Retrieve, re-rank and compress context, then ground the answer with citations and refusals.
- Top-k + re-ranking
- Context compression
- Grounded citations
- 07
Orchestration & tools
Wrap the model in typed tools, structured output and agent loops with hard limits.
- Function / tool calling
- Structured output schemas
- Step & cost limits
- 08
Evals & guardrails
Score every change against the golden set, and keep prompt injection, PII and jailbreaks out.
- Offline evals
- LLM-as-judge
- Injection & PII filters
- 09
Ship & observe
Stream responses, rate limit per user, trace every call and watch quality, latency and spend in production.
- Streaming UX
- Rate limiting
- Tracing & cost dashboards
06 · Side builds