Senior AI Engineer with RAG and Agentic architectures İş İlanı

We are looking for a Senior AI Engineer to design and deliver LLM-powered applications and agentic systems that solve real problems for our users and teams. You'll work across the full applied-AI stack — from retrieval-augmented generation (RAG) pipelines and prompt design to multi-step agents that reason, use tools, and automate end-to-end workflows. This is a senior builder's role. You'll take challenging use cases from an ambiguous problem to production: choosing the right models, implementing agentic architectures, wiring up retrieval and tooling, defining how we evaluate quality, and shipping reliable applications at scale. Responsibilities Design and deliver production LLM applications — chat, copilots, assistants, and autonomous workflows — from concept to scale Build and implement RAG pipelines: chunking, embeddings, vector search, reranking, and grounding to reduce hallucination and improve relevance Develop agentic architectures: multi-step reasoning, tool/function calling, planning, memory, and multi-agent orchestration Contribute to the automation of business and engineering workflows using agentic AI and workflow automation Define prompt and context strategies; build evaluation harnesses and uphold quality, latency, and cost standards Integrate LLMs with internal data, APIs, and tools via connectors, function calling, and structured outputs Implement guardrails, safety, and observability for AI systems (tracing, evals, monitoring for quality and drift) Collaborate with product, data, and platform teams to turn ambiguous problems into shipped AI features Share knowledge with fellow engineers and participate in design reviews Requirements 3+ years in software or ML engineering, including recent, hands-on experience building and shipping applications with LLMs Proven experience delivering applied-AI systems end to end Proficiency in Python at an advanced level Hands-on background in building RAG systems — embeddings, retrieval, and reranking with vector databases (Pinecone, Qdrant, Milvus, or pgvector) Expertise in LLM APIs and orchestration frameworks (OpenAI, Anthropic, LangChain, or LlamaIndex) Skills in designing agentic architectures — tool use, function calling, planning loops, and agent orchestration in production Competency in automating workflows with agentic AI or workflow-automation tooling Knowledge of prompt engineering and structured/JSON output techniques Capability to design evaluations and reason about LLM quality, cost, and latency trade-offs at scale Excellent command of written and spoken English (B2+ level) Nice to have Familiarity with multi-agent frameworks (LangGraph, CrewAI, or AutoGen) Experience with fine-tuning, adapters (LoRA), or model distillation Understanding of MLOps/LLMOps — deployment, versioning, and monitoring of AI systems, including model serving and inference optimization Knowledge of AI safety, guardrails, and evaluation frameworks (Ragas, LangSmith, or promptfoo) Background in cloud platforms (AWS, GCP, or Azure) and containerization (Docker, Kubernetes) We offer CONTINUOUS UPSKILLING, LEARNING & DEVELOPMENT Diversity of tasks and projects Assessment center for objective review of competency level Personal development plan Mentoring programs and leadership development Certification and professional development support Access to learning platforms including more than 2,500 internal courses English courses taught by certified teachers CORPORATE BENEFITS Extra leave days Referral bonuses COMPENSATION PACKAGE Competitive compensation paid in USD Regular salary and performance reviews MEDICAL & HEALTHCARE Private health insurance Well-being events WORKING ENVIRONMENT Recreation areas and kitchens Tea, coffee and snacks Sports equipment and game consoles IT Equipment Microsoft’s Software Assurance Home Use Program (HUP) Please note that our Talent Attraction Team reviews applications and CVs submitted in English. EPAM is global leader in AI transformation engineering and integrated consulting, serving Forbes Global 2000 companies and ambitious startups. With over thirty years of expertise in custom software, product and platform engineering, we empower our clients to become AI-Native enterprises, driving measurable value from innovation and digital investments.