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. Brightstar has approximately 6,000 employees. For more information, please visit www.brightstarlottery.com.
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, 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 LangChain, HuggingFace Transformers, or OpenAI API.
Design prompt templates and implement retrieval strategies to increase answer precision and factuality.
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 GenAI.
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., LangGraph, AutoGen, CrewAI).
Hands-on expertise with Generative AI (LLMs, prompt engineering, fine-tuning).
Proficiency in Python and familiarity with deep learning/NLP libraries (LangChain, PyTorch, TensorFlow, HuggingFace 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 (eg: 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) for event-driven agent triggers.
Experience implementing RAG using Azure AI Search (vector, semantic, and hybrid search) or Databricks One.
Qualifications:
Bachelor's degree in computer science, engineering, mathematics, physics, economics, or a related field.
Minimum 3 years of experience in technical delivery roles, with experience in moving products, internal tools, or workflow systems beyond proof-of-concept stage.
Minimum 3 years of hands-on experience in one or more of the following areas: backend services, APIs and integrations, full-stack delivery, data pipelines, model-serving or machine learning workflows, or agentic orchestration systems.
Strong coding ability in Python plus one complementary engineering surface such as TypeScript or JavaScript, front-end delivery, cloud or platform engineering, or infrastructure automation.
Sound engineering judgment around enterprise hardening and evaluation, including experience with several of the following: authentication, role-based access control (RBAC), observability, security, release discipline, regression testing, or experiment frameworks for AI, machine learning, or agentic workflows.
#LI-YG1 #Hybryd
This information reflects Brightstar Lottery’s good-faith estimate of compensation for this position at the time of publication. Actual compensation will be determined in accordance with the applicable collective agreement and internal compensation policies. Factors that we consider include, but are not limited to, role, career level, performance, skills, level of expertise, experience, location, market competitiveness, and internal equity. The estimated starting compensation range is € 41.586 – € 63.979. The Company will comply with all local pay requirements and collective bargaining agreements, where applicable.
In addition to base salary, employees may be eligible for an annual performance-based bonus with a set target (% of salary) aligned with individual and company performance. Eligibility and payout may be prorated based on date of hire and service during the performance period or sales commission.
Other Benefits Offered:
Paid time off
Health insurance, life insurance, and disability insurance
Opportunities to volunteer your time to company-driven initiatives, employee networks or organizations of your choice
Variety of well-being programs
Retirement benefits including Supplementary Pension
Additional benefits available depending on the seniority of the role
Mobility Benefits
Meal Vouchers
This position is governed by collective bargaining agreement, which regulates the main terms and conditions of employment, including minimum salary levels, classification, working hours, leave, and other employment rights.
Our hiring process is conducted in line with principles of pay transparency and equal pay. We do not request or rely on candidates’ salary history during the selection process.
All Brightstar employees have a role in information security. Annual training will be assigned and required as appropriate.
