AI Engineer

Location:
Roma, IT, 00154 Requisition ID: 19460 Brightstar is an innovative, forward-thinking global leader in lottery that builds on our renowned expertise in delivering secure technology and producing reliable, comprehensive solutions for our customers. As a premier pure play global lottery company, our best-in-class lottery operations, retail and digital solutions, and award-winning lottery games enable our customers to achieve their goals, fulfill player needs and distribute meaningful benefits to communities. We have a longstanding commitment to Responsible Gaming (RG) that is engrained within our core business and the products

we offer
to customers and players worldwide. Brightstar has a well-established local presence and is a trusted partner to governments and regulators around the world, creating value by adhering to the highest standards of service, integrity, and responsibility. For more information, please visit Role Overview We develop and operate a highly data-intensive platform offering analytics, machine learning and agentic AI on cloud environments. As part of continued investment, we are building a cutting-edge AI-native decision-system. As an Agentic AI Developer, you will be responsible for designing and developing advanced agentic AI solutions capable of autonomous planning, reasoning, and task execution. Developing and integrating Retrieval-Augmented Generation (RAG) pipelines to enhance agents' reasoning capabilities with external knowledge sources, and multi‐agent communication protocols to enable seamless collaboration and coordination among agents. Collaboration will be key, as you work cross‐functionally with Data Scientists, ML Engineers, and Product Teams to deliver robust, end‐to‐end AI solutions. You will work at the intersection of cutting‐edge AI, Data Science and modern software engineering, implementing multi‐agent frameworks and orchestrating interactions.

Key Responsibilities
Design and implement Agentic AI architectures for enterprise workflows. Integrate Generative AI capabilities (LLMs, multimodal models) into client solutions. Deliver end‐to‐end AI solutions from ideation to production deployment. Build, fine‐tune, and evaluate LLM‐based Q&A models using frameworks like Lang Chain, Hugging Face Transformers, or Open AI API. Collaborate with product managers to translate user requirements into technical features. Participate in error analysis, iterative model improvement, and performance tuning. Engage with clients to identify high‐value AI use cases and define business benefits. Conduct workshops and assessments to align AI strategies with organisational goals. Provide thought leadership on AI adoption and emerging trends. Develop reusable frameworks and accelerators for Agentic AI and Gen AI. Ensure compliance with AI ethics, security, and governance standards. Stay on top of industry developments in Agentic AI, autonomous agents, and LLM ecosystems. Orchestrate complex multi‐agent workflows to handle tasks requiring planning, reasoning, and tool use. Extend Microsoft 365 Copilot by building custom plugins and declarative agents within Microsoft Copilot Studio to surface enterprise data in Teams and Office apps. Operationalize AI solutions using Microsoft AI Foundry for model catalog management, Prompt Flow evaluation, and lifecycle governance. Architect scalable deployment patterns for agents using Azure Container Apps or Azure Functions, ensuring low‐latency responses and cost‐effective scaling. Platforms and interfaces Own data flows, APIs, services, model‐serving surfaces, front‐end and desktop application surfaces, continuous integration and continuous delivery (CI/CD), and demo hardening. Build the systems that make quantitative work feel polished, reliable, and enterprise‐ready for expert users and client stakeholders. Agent‐assisted systems Own the agentic harness layer — evaluation frameworks, reviewer loops, control‐plane behavior, orchestration, and tool integration — that applications and MCPs wrap around. Design opinionated harnesses that expose through MCP or similar integration patterns without overfitting to one vendor or one moment in the tooling market. Requirements Strong experience in Agentic AI frameworks (e.g., Lang Graph, Auto Gen, Crew AI). Hands‐on expertise with Generative AI (LLMs, prompt engineering, fine‐tuning). Proficiency in Python and familiarity with deep learning/NLP libraries (Lang Chain, Py Torch, Tensor Flow, Hugging Face Transformers). Experience with building Q&A systems and retrieval‐augmented generation pipelines. Knowledge of vector databases or semantic search concepts. Familiarity with cloud AI platforms (e.g., Azure AI Foundry). Knowledge of MLOps practices and deployment pipelines. Experience with Git, collaborative development workflows, and cloud infrastructure. Proficiency in deploying AI workloads to Azure Container Apps (ACA), Azure Kubernetes Service (AKS), Databricks Apps or serverless functions (Azure Functions)