Senior AI Engineer

Deviniti is a company based in Wrocław specializing in developing production-grade generative AI solutions for enterprise clients in regulated industries. The team consists of 14 people including AI Architects, ML Engineers/Data Scientists, Python Developers, ML Ops, and FullStack Developers. The Application Development unit focuses on custom generative AI solutions involving design, development, client verification, and production deployment, emphasizing data privacy, security, auditability, and model quality control.

• Minimum 4 years of experience developing production-grade solutions in Python.
• Practical knowledge of FastAPI and asynchronous programming.
• Experience working with LLM and generative AI solutions, especially RAG, vector databases, retrievers, hybrid search, and prompt engineering.
• Familiarity with LangChain or LlamaIndex.
• Experience with one of: Azure OpenAI, Amazon Bedrock, or OpenAI API.
• Knowledge of PostgreSQL with pgvector or Qdrant.
• Basic knowledge of Elasticsearch or OpenSearch.
• Experience with Docker and basic knowledge of Kubernetes or OpenShift.
• Experience with AWS or Azure cloud platforms.
• Practice with CI/CD, unit and integration tests, Git, and code review.
• Knowledge of Linux, networking basics, and security including OIDC/OAuth2, KMS/Secrets.
• Experience in observability: logs, metrics, and tracing.
• Skills in profiling and optimizing latency and costs of AI solutions.
• Polish and English language skills at least B2 level.
• Awareness of data security and GDPR requirements in generative AI contexts.

Nice to have:

• Knowledge of LangGraph, RAGFlow, Dify, n8n or Temporal.
• Experience with Ragas or Evals, prompt-CI, A/B testing.
• Knowledge of OpenShift, Helm, Terraform, Keycloak.
• Experience with RabbitMQ, gRPC, event-driven patterns.
• Familiarity with MLflow or W&B, including basic fine-tuning and multimodal capabilities.
• Skills in TypeScript, Next.js, or Streamlit.
• Production experience with Elasticsearch or OpenSearch.
• Experience in regulated sector projects.
• Knowledge of ISO 42001, ISO 27001/27017 standards and basic SOC 2 concepts.

• Design and implement generative AI applications, including Retrieval-Augmented Generation (RAG) solutions.
• Develop APIs and serving layers using FastAPI.
• Build pipelines for chunking, embeddings, indexing, hybrid search, and reranking.
• Evaluate solution quality offline and online, prepare regression tests for RAG, and monitor quality and costs.
• Implement guardrails such as content filtration, prompt security, and access policies.
• Integrate with client systems via REST, gRPC, ESB, or EventBus.
• Perform monitoring and observability activities, including logging, tracing, and metrics with OpenTelemetry.
• Optimize performance, model usage costs, and latency.
• Document technical decisions through ADRs and model cards.
• Collaborate with frontend, backend, DevOps, business analysis, and project management teams.
• Participate in client discussions for solution verification and project adjustments.
• Support sales during presale stages.

The administrator of personal data is Deviniti with headquarters in Wrocław, ul. Sudecka 153. Personal data is processed for recruitment purposes with rights to access, rectify, delete or restrict data and to object to processing. Providing data required by labor law is mandatory; other data is voluntary. Data retention follows legal and consent timelines. Candidates have the right to withdraw consent anytime. Data recipients include Just Join IT and related entities processing recruitment data. Deviniti has implemented whistleblower protection procedures compliant with applicable laws.