Senior AI Fullstack Engineer at Levi9 Ukraine – UA | Jobyssey

🔹Who we are!
At Levi9, we are passionate about what we do. We love our work, and together as a team, we are smarter and stronger. We are looking for skilled team players who make change happen. Are you one of these players?

🔹About the project
Our client is a leading media company in Western Europe, delivering high-quality content across newspapers, magazines, radio, TV, websites, apps, and podcasts. Their brands reach millions of people every day and help shape the media landscape through independent and trusted journalism.
We are developing AI capabilities for the digital advertising domain, helping improve campaign management, analytics, operational efficiency, and user engagement. The team will create a range of AI-driven solutions that address real business needs and evolve alongside the platform and market.
These solutions may combine agentic workflows, large language models, enterprise data, and backend integrations. They are designed with security, scalability, automated evaluation, observability, and cost efficiency in mind and are deployed as modular services on AWS.

🔹 Your role
As a Senior AI Engineer, you will design, build, and operate production-grade AI agents and the services around them. You will own features end to end — from clarifying business needs and selecting an approach to implementation, evaluation, deployment, monitoring, and continuous improvement.
You will work closely with backend, frontend, platform, and product specialists to apply AI to complex advertising workflows and business challenges. You will also help establish engineering practices for reliable LLM-based systems, including prompt and agent evaluation, security guardrails, observability, and cost-efficient model usage.
🔹Responsibilities
Design and develop specialized AI agents for advertising and analytics use cases.
Build reliable agent tools and integrations using HTTP APIs, OpenAPI, and MCP.
Develop and improve Text-to-SQL workflows with strict read-only access and query-safety controls.
Implement dynamic prompts, runtime context, memory, intent routing, and streaming responses.
Define automated evaluation suites for functionality, quality, security, and regression testing.
Introduce safeguards against prompt injection, jailbreak attempts, unsafe tool use, and data leakage.
Trace, monitor, and troubleshoot agent interactions in development and production.
Build containerized services and support their deployment and operation on AWS.
Collaborate with stakeholders to refine requirements and translate them into maintainable technical solutions.
Contribute to architecture decisions, code reviews, engineering standards, and knowledge sharing within the team.
Explore new models, frameworks, and agent patterns and apply them where they deliver measurable value.

🔹Technology stack
AI and agents: Mastra, large language models, prompt engineering, tool calling, agent memory
Protocols and integration: Model Context Protocol (MCP), OpenAPI, HTTP, REST, JSON
Backend: TypeScript, Node.js 20+
Data: PostgreSQL, Text-to-SQL, Zod
Frontend: React, TypeScript, Vite
Quality and evaluation: Promptfoo, Vitest, automated functional and security evaluations
Observability: Langfuse, distributed tracing, structured logging
Cloud and delivery: AWS, ECS, Docker, CI/CD

🔹Requirements
3+ years of professional software engineering experience, including hands-on work with LLM-powered or agentic applications.
Strong TypeScript and Node.js skills and practical experience building backend services and APIs.
Experience designing LLM workflows involving tool calling, structured outputs, retrieval, routing, or multi-agent coordination.
Hands-on experience designing and implementing Retrieval-Augmented Generation (RAG) pipelines, including document processing, embeddings, vector search, and retrieval evaluation.
Practical understanding of prompt engineering and systematic evaluation of non-deterministic systems.
Experience integrating services through REST APIs and working with OpenAPI specifications.
Knowledge of identity, authorization, least-privilege access, and secure context propagation.
Understanding of AI security risks, including prompt injection, excessive agency, unsafe tool execution, and sensitive-data exposure.
Experience with automated testing, Docker, CI/CD, monitoring, and production support.
Hands-on cloud experience, preferably with AWS and containerized workloads.
Strong stakeholder communication skills, including the ability to clarify business needs, explain technical solutions and trade-offs, and identify the most effective way to achieve business goals.
Strong software design, problem-solving, and ownership skills.
Curiosity and enthusiasm for emerging AI capabilities, with a pragmatic approach to evaluating and applying them to real business needs.
Upper-intermediate or higher English proficiency.

🔹Nice to have
Production experience with Mastra or another modern AI agent framework.
Hands-on experience implementing or consuming MCP servers and clients.
Experience building and evaluating Text-to-SQL systems.
Familiarity with Promptfoo, Langfuse, or similar AI evaluation and observability platforms.
Experience with OpenAI-compatible APIs and comparing or migrating between model providers.
Knowledge of React and streaming conversational interfaces.
Familiarity with AWS ECS, and cost optimization for AI workloads.
Experience working with advertising technology, analytics platforms, or high-volume data products.

🔹9 reasons to join us:
Today we're working with the technology of tomorrow.
We don't wait for a change. We are the change.
We're experts in creating experts (Levi9 academy, Lead9 program for leaders).
No micromanagement. We are free birds with a clear understanding of what the high performance is!
Learning in Levi9 never stops (unlimited Udemy for business, meetups, English courses, Professional trainings).
Here you can train your body and mind.
We've gathered the best locations - comfortable, cozy and pet-friendly offices in Kyiv (5 minutes from Olimpiyska subway station) and Lviv (5 minutes from the Opera Theatre) with regular offline internal events
We have master's degree in work-life balance.
We are actively supporting Ukraine with constant donations and volunteering

Explore more jobs