About the Team
We are Quadcode, a fintech company excelling in financial brokerage activities and delivering advanced financial products to our global clientele. Our flagship product, an internal trading platform, is offered as a Software-as-a-Service (SaaS) solution to other brokers.
We're looking for an AI Engineer to join our AI Platform team, which builds and operates the generative AI pipelines powering our products, many of which involve financial and trading data, terminology, and specifications. This is a backfill role on a small, high-impact team, so you'll have real ownership from day one.
You'll work across our broader AI-platform portfolio: prompting and prompt evaluation, agentic systems, MCP tooling, and evaluation pipelines: writing backend Python on top of LLM providers like OpenAI, Anthropic, and xAI. This is an applied engineering role focused on shipping and maintaining production LLM systems, not a model-training or research position, prior training/research experience is not required.
This role is hybrid, with an expectation of at least 3–4 days per week in our Limassol office.
Tasks:Design, build, and maintain LLM-powered features and pipelines for our generative AI products, including those handling financial and trading contentDevelop and improve agentic systems and multi-agent workflows, including integrations using MCP (Model Context Protocol) servers/toolingWrite production-grade backend Python services on top of LLM provider APIs (OpenAI, Anthropic, xAI) and/or on-premise LLMsBuild and iterate on prompt engineering and prompt-evaluation pipelines, including structured output generation and validationTake prototypes and proofs-of-concept and turn them into reliable, maintainable production servicesMonitor and manage LLM cost, latency, and reliability trade-offs across pipelinesCollaborate closely with a small, senior AI Platform team on architecture, code review, and technical direction.Requirements:Production LLM experience: You've shipped real, live features using OpenAI and/or Anthropic APIs, or with on-premise/open-weight LLMs, not just experimentation or side projectsMulti-agent & MCP experience: Hands-on experience building multi-agent frameworks and working with MCP serversStrong Python & software engineering fundamentals (2–4+ years): REST APIs, SQL, Docker, git, and concurrency, with a demonstrated ability to take projects from prototype to maintainable production servicePractical LLM engineering skills: Comfortable working with structured outputs, prompt evaluation, and cost-aware design decisions in real systemsEnglish fluency (written and verbal).Nice to Have:Prior experience working in fintechFinance/markets knowledge: tickers, earnings, technical indicators, forex/CFDsQuant backgroundExperience with model fine-tuning/training, RAG, or LLM optimization techniquesTrack record in Kaggle competitions focused on LLMs or multi-agent frameworksGPU infrastructure experience: provisioning GPU compute or serving open-weight models (e.g., Hugging Face inference)Experience with Kubernetes, Kubeflow, or AirflowData analysis and statistics experienceExperience with cloud platforms (AWS, GCP, etc.)Russian language proficiency.
We know the combination of production LLM engineering and finance-domain knowledge is rare. If you're a strong AI/LLM engineer without a finance background, we still want to hear from you.
What we offer:Hybrid work model in our brand-new office in LimassolHealth insurance and mental health services13th salary and 21 vacation days per yearCatered lunches in the officeTuition reimbursement (kindergartens/schools)Training and Development opportunitiesOnsite GymCorporate events and workshopsBonuses for special events (e.g., child's birth).With a team of 500+ creative and tech-driven people across the globe, Quadcode is the place where ideas turn into impact, where every project is a chance to grow, innovate, and make your mark.
Join us today, and let's shape the future of fintech together!
Note: All applications will be treated with strict confidence. We thank all applicants for their interest, however, only those candidates selected for interviews will be contacted.
#LI-LR1