Backend Engineer
Senior AI Engineer — LLM & Agentic Systems
Warsaw
Apply on the employer's siteRole description
All open positions
AIRemote / Hybrid 4 min read
Senior AI Engineer — LLM & Agentic Systems
Production-grade LLM systems, RAG pipelines and agentic workflows for enterprise use — end-to-end.
Employment Full-time
Location Warsaw / Remote (EU)
Work mode Remote-first / Hybrid in Warsaw
Job Overview
We are building an AI Engineering team focused on production-grade LLM systems. Not another chatbot demo, not a playground that never reaches users.
Our work sits between software engineering, applied AI, data systems, cloud infrastructure, evaluation, security and product thinking. We build LLM-powered systems that run on real business data, support real users and are designed to be monitored, evaluated, secured and maintained in production.
We need a Senior or Staff-level AI Engineer who can own the full path from problem understanding to production delivery. That means architecture, model and tooling choices, implementation and shipping.
This role is a strong fit for someone who likes pragmatic engineering, understands the limits of current LLMs and knows that a reliable AI system is much more than a prompt and an API call.
What You Will Build
- RAG systems for unstructured and semi-structured data: documents, reports, tickets, knowledge bases and CRM/ERP records.
- Agentic workflows with tool calling, task planning, permissions, auditability and human-in-the-loop control.
- AI assistants for operational, analytical and product teams.
- Evaluation frameworks for answer quality, retrieval quality, prompt regressions and model behavior.
- Integrations between LLMs, internal APIs, data warehouses and business systems.
- Guardrails for sensitive data, hallucination control, prompt injection defense and traceability.
- Cost, latency, model selection, fallback and routing optimization.
Responsibilities
- Design and develop production-grade LLM systems and agentic workflows.
- Build and optimize RAG pipelines: chunking, embeddings, retrieval, reranking, grounding and citation strategies.
- Design agent architectures: tools, memory, planning, routing, retries, timeouts, permission models and escalation paths.
- Integrate commercial and open-source models with applications, APIs and data systems.
- Create evaluation frameworks: offline evaluation, regression tests, test sets, LLM-as-judge and quality gates.
- Implement observability for AI systems: tracing, cost monitoring, latency monitoring, token usage, fallbacks and alerts.
- Work on AI security and governance: PII redaction, data access control, prompt injection defense and audit logs.
- Collaborate closely with Product, Data Science, Backend, Security and Cloud Engineering teams.
- Make technology decisions: when to use LangGraph, when to build simpler orchestration, when to use RAG, classic search, fine-tuning or a hybrid approach.
- Help shape AI Engineering standards: architectural patterns, repository templates, review checklists, documentation and delivery practices.
Technology Stack
- Core engineering
Python, FastAPI, Pydantic, SQL, PostgreSQL, Docker, GitHub Actions
- LLM / AI
OpenAI / Azure OpenAI, Anthropic, open-source LLMs, Hugging Face, embeddings, rerankers, structured outputs, function calling
- Frameworks
LangGraph, LangChain, LlamaIndex. Experience building lightweight custom orchestration instead of relying on frameworks by default is a strong plus
- RAG / search
pgvector, OpenSearch / Elasticsearch, Qdrant, Pinecone, Weaviate, hybrid search, reranking, metadata filtering
- MLOps / LLMOps
MLflow, Langfuse, LangSmith, OpenTelemetry, Prometheus, Grafana, test sets, golden datasets, evaluation pipelines
- Cloud / infrastructure
AWS, Azure or GCP, Kubernetes, Terraform, serverless, CI/CD
- Data
Snowflake, BigQuery, Databricks, Airflow, dbt or similar tools are a plus
Requirements
- At least 5 years of commercial experience in software engineering, AI/ML engineering, data engineering or backend engineering.
- Strong Python skills and solid engineering fundamentals: testing, code review, modular design, clean architecture and CI/CD.
- Practical experience with LLM systems, RAG or agentic workflows that goes beyond local demos or notebooks.
- Experience with LangGraph, LangChain, LlamaIndex or the ability to consciously design a simpler alternative.
- Hands-on experience with vector databases, semantic search or hybrid search.
- Good SQL skills and experience working with production data.
- Ability to design systems that are reliable, observable and maintainable in production.
- Comfortable communication in English in an international environment.
- Ownership mindset. You care about the outcome, not just the ticket.
Nice to Have
- Experience in regulated environments such as healthcare, fintech, insurance, legal, pharma or enterprise.
- Experience with MLOps / LLMOps and evaluation pipelines.
- Hands-on experience with Kubernetes, Terraform and public cloud platforms.
- Experience with model serving tools such as vLLM, TGI, Triton or similar.
- Understanding of LLM security risks: prompt injection, data leakage, access control and auditability.
- Experience with large document collections, knowledge graphs or GraphRAG.
- Previous experience as a Tech Lead, Staff Engineer or architecture owner.
What We Offer
- Clear cooperation terms and transparent workload expectations.
- Remote-first setup and flexible working hours.
- Budget for conferences, training and AI/cloud certifications.
- Equipment budget suitable for serious engineering work.
- Access to paid AI tools, test environments and experimentation budget.
- Time for research, prototyping and validating new approaches.
- Private medical care and benefits package.
- No unnecessarily long recruitment process.
- Participation in decisions about the AI roadmap, engineering standards and team direction.
Who We're Looking For
We are looking for someone who understands that a production AI system is not just a prompt and an API call.
It is data, retrieval, evaluation, monitoring, cost, security, UX, deployment and long-term maintenance.
If you enjoy building AI systems that work beyond demo day, we should talk.
Valora
Recruitment Process
A short, focused process designed for senior practitioners.
- 1 Application
You submit your CV and consent. That's it.
- 2 CV Review
We review your background within a few business days.
- 3 Technical Interview
A focused conversation about real engineering problems.
- 4 Client Interview
Only when the role requires it, never a repeat of step 3.
- 5 Offer
Clear terms, transparent conditions, quick decision.
Apply for this role
Warsaw / Remote (EU)
Remote-first / Hybrid in Warsaw
Apply for this role
PythonFastAPILangGraphLangChainLlamaIndexpgvectorOpenAIAnthropicKubernetes
Apply for this role
This is a saved copy of a posting published elsewhere. Postings get taken down without notice — check the employer's site before applying. mentors.coach is not the hiring party.