Full-stack Engineer

Full-stack Engineer Submit your resume and join our team! EverHelp is a team of professionals developing outsourcing support services for products all over the world at a dynamic tempo. The project launched in 2021, and last year we experienced a yearly growth rate of +82%. Since February 2022, we've managed to preserve all our workplaces and, most importantly, we continue to strengthen our team while maintaining an employee satisfaction rate of over 90%. In 2026, EverHelp received 9 international workplace and employer awards. The company ranked #1 in the Top 30 Inspiring Workplaces in Europe, was named the #1 employer among European companies with 500–4,999 employees, and earned Best-in-Class recognition across 6 categories: Culture & Purpose, Leadership, Inclusion & Belonging, Wellbeing, Employee Voice, and Employee Experience. EverHelp was also named one of the 28 Best Places to Work by Business Intelligence Group — a global award recognizing companies with strong workplace cultures and outstanding employee experience. Our statement: We're against war and unjustified aggression, We evacuated all the teammates and their families from the frontline, helped them find a new place to live and provided financial support, We provided the team with charging stations, We continue to work together for the future of Ukraine. About the role EverHelp is building a product portfolio from scratch: Evly (AI customer service), analytics, a voice agent, and Zendesk Migration. We're looking for a Full-stack Engineer (Agentic Systems) to own end-to-end product development — from the client-facing web layer to backend business logic, including agentic engineering, highload system design, and shipping AI features against real business and product requirements. This is an engineer who turns a product idea into a working, scalable feature by closing the whole stack independently, not just one layer of it. Your future responsibilities include: End-to-end feature development : frontend, backend business logic, APIs, and integrations — owning the full stack independently System design for highload systems: scalability, queues, caching, and resilience under load Agentic engineering: AI agents, Copilot-style flows, and LLM orchestration (tool/function calling, prompt engineering, memory, fallback logic) RAG systems: embeddings, vector search, and building/optimizing retrieval pipelines Working with pre-deployed self-hosted models and integrating LLMs into the product AI security : data protection, prompt injection defense, access control, handling sensitive data LLM evaluation & optimization: quality metrics, observability (logging, tracing, latency/cost), and reducing costs without sacrificing quality Data work: database modeling, SQL and MongoDB (document/NoSQL structures), and preparing data for analytics and AI flows Independently deploying features on top of existing infrastructure (via pre-built Helm chart templates) Code review, test coverage, and maintaining the technical health of the codebase Maintaining up-to-date technical documentation : architecture, API contracts, AI flows Clarifying requirements, estimating and decomposing tasks, and raising technical risks early (in collaboration with design) Needed experience & skills: Frontend React + Next.js, TypeScript/JavaScript, Vite, TailwindCSS — able to build client applications independently Understanding of UI state, API integration, and basic attention to UX and interface performance Backend Node.js and Python — business logic, REST APIs, WebSockets, integrations Database design and hands-on work with PostgreSQL and MongoDB; confident SQL Queues and caching: RabbitMQ, Redis; object storage (S3) Containerization with Docker; deploying services via existing Helm chart templates AI / LLM Engineering Agentic engineering: building AI agents with LangChain, LangGraph, AutoGen, CrewAI, or similar Tool/function calling, prompt engineering, memory and fallback logic RAG systems: embeddings & vector search; vector DBs such as Qdrant, pgvector, or Pinecone Experience working with pre-deployed self-hosted models LLM evaluation & optimization: quality, latency, and cost metrics; AI solution observability Understanding of AI security: prompt injection, handling sensitive data General Systems design and understanding of highload architecture (scalability, resilience under load) Observability: instrumenting services with logs and metrics; familiarity with the Grafana/Loki/Prometheus stack Code under test: unit, integration, and load testing; code review discipline; clean, documented codebase English B2+ (technical documentation, external integrations, written communication) 2–3 years of experience as a Full- stack developer, plus at least 1 year in AI engineering Nice to have skills: Experience with the customer support/CX domain (Zendesk, Intercom, Freshdesk — as a user or via integrations) Voice AI experience: voice agents, STT/TTS, real-time audio pipelines, LiveKit, SIP telephony MLOps / deploying and maintaining self-hosted models (fine-tuning, quantization, GPU inference) Experience fine-tuning or adapting LLMs for a specific domain Experience launching a product from 0→1, working at startup pace Building analytics dashboards and working with product metrics (Amplitude or similar), BigQuery Experience with orchestration systems (Temporal or similar) Work with EverHelp is about: Opportunity to join the Ukrainian company and help the Ukrainian economy, 20+ vacation days and unlimited sick leaves, corporate doctor & psychologist, Medical insurance and +10 Healthcare leaves for mental & physical recovery, Systematic learning & development inside a team and on an individual level: compensation for professional webinars, courses, conferences, etc., Ability to work from Lviv & Kyiv offices with free breakfasts and lunches, or fully remotely , Team of professionals ready for knowledge sharing, internal professional communities, team buildings, Corporate discounts for gym membership and other services like medical services, language schools, etc. Department Product and Analytics Department Remote status Fully Remote