Senior Data/ai Engineer

AT&T AT&T · Telecom · CZE:642:Brno +2

Senior Data/AI Engineer to build, integrate, and deploy enterprise-grade AI systems focused on LLMs, prompt engineering, and context management. The role involves designing context-rich AI solutions, crafting prompts, integrating agentic frameworks like LangGraph, and building data pipelines for scalable deployment. Responsibilities include developing and deploying context-driven AI solutions using LLMs, RAG, and AI agents to automate workflows, enable autonomous agents, and drive productivity.

What you'd actually do

  1. Design, implement, and optimize prompts for LLMs to achieve accurate and contextually relevant outputs across various use cases.
  2. Develop and maintain context management strategies, including session memory and retrieval-augmented generation, to improve agent performance.
  3. Integrate and support fine-tuning of LLMs in Python-based applications, leveraging APIs and pipelines for scalable deployment.
  4. Build, test, and maintain conversational and agent workflows using frameworks like LangGraph for multi-agent or multi-step solutions.
  5. Develop and maintain backend services, APIs, and possibly front-end interfaces to deliver end-to-end AI applications.

Skills

Required

  • fullstack Python development (FastAPI, Flask; SQL/NoSQL databases)
  • prompt engineering for LLMs (OpenAI, Anthropic, open-source LLMs)
  • context engineering, including session management, vector search, and knowledge retrieval
  • integrating AI agents and LLMs into production environments
  • agentic/conversational frameworks such as LangGraph
  • Retrieval-Augmented Generation (RAG) concepts and multi-agent orchestration
  • cloud infrastructure, containerization (Docker), and CI/CD
  • API development patterns (e.g., HTTP/REST, GraphQL) and microservice architecture

Nice to have

  • data engineering technologies (e.g., Spark, Kafka, Databricks, Snowflake) and related languages (e.g., Java, Scala, SQL)
  • AI-powered IDEs and tools such as Visual Studio, GitHub Copilot, Windsurf, Cursor
  • fine-tuning language models
  • statistics and machine learning (probability, hypothesis testing, experimental design, bias/variance, regularization)
  • building and validating predictive models end-to-end (feature engineering, training, cross-validation, model selection, error analysis) with tools like scikit-learn and XGBoost/LightGBM/CatBoost
  • supporting evaluation strategies: offline metrics, experimentation (A/B testing), guardrails, and monitoring for drift/quality
  • LLM evaluation frameworks (automated and human-in-the-loop), measuring generation quality (success, relevance, factuality, safety), and using evaluation patterns such as LLM-as-a-judge
  • RAG-focused data science: retrieval evaluation (recall@k, MRR, nDCG), chunking/embedding tradeoffs, query rewriting, reranking, grounding

What the JD emphasized

  • fullstack Python
  • modern data and AI engineering technology stack
  • LLMs
  • prompt engineering
  • context management
  • agentic frameworks like LangGraph
  • autonomous AI agents
  • Retrieval-Augmented Generation (RAG)
  • modern AI agents
  • conversational and agent workflows
  • multi-agent or multi-step solutions
  • backend services, APIs, and possibly front-end interfaces
  • prompt experiments
  • production-grade Python applications
  • agent deployment
  • prompt and context strategies
  • prompt engineering for LLMs
  • context engineering
  • session management
  • vector search
  • knowledge retrieval
  • AI agents and LLMs into production environments
  • agentic/conversational frameworks such as LangGraph
  • Retrieval-Augmented Generation (RAG) concepts
  • multi-agent orchestration
  • fine-tuning language models

Other signals

  • building enterprise-grade AI systems
  • LLMs, prompt engineering, and context management
  • integrating agentic frameworks like LangGraph
  • autonomous AI agents