• Design, train, evaluate, and deploy machine learning systems that power real-time voice experiences, including ASR, speech understanding, turn detection, text to speech, speech to speech, classification, entity extraction, summarization, and structured insight generation. • Improve the quality of voice AI systems through error analysis, data curation, metric design, benchmarking, and iterative model improvement, with a strong focus on real-world performance. • Build evaluation frameworks for complex voice and agentic systems, measuring metrics such as accuracy, robustness, latency, faithfulness, naturalness, professionalism, task completion, and cost. • Diagnose and mitigate failure modes across the voice stack, including transcription errors, hallucinations, retrieval failures, tool misuse, prompt brittleness, context drift, and multi-step reasoning breakdowns. • Design and optimize low-latency ML workflows for live conversations, balancing model quality with system responsiveness, scalability, and reliability. • Partner with platform and backend engineers to productionize real-time inference, streaming pipelines, quality monitoring, and continuous model iteration. • Collaborate cross-functionally with product, design, frontend, and backend teams to integrate voice intelligence seamlessly into Cresta’s platform. • Establish best practices for offline evaluation, online experimentation, model validation, observability, and ongoing quality monitoring in production. • Mentor engineers, contribute to technical strategy, and help shape the roadmap for Cresta’s voice AI systems.

Role Description We have to fill multiple positions. Depending on your previous work experience (years in the job market), qualification (Bachelor, Master, Ph.D., Post-doc) you might be offered two types of position: - No work experience, and a Master / Bachelor → We offer you an Internship - At least 2 years of work experience OR a Ph.D. / Postdoc → We offer you a Full-Time Position What You’ll Do? Depending on your profile and interests, you could be placed on one of two flagship tracks: - Track 1. AI for Financial Services: Work on Machine Learning solutions for one of the most data-rich industries in the world. Problems you might tackle include: - Prediction models - Document analysis with ML - Fine-tuning LLMs for conversational data interfaces - Extracting actionable insights from large-scale datasets - Track 2. Computer Vision for Pharma: Contribute to Computer Vision pipelines deployed in pharmaceutical environments, focusing on: - Object tracking - Behaviour understanding - Solving complex real-world problems with nothing but a camera and a well-trained model Both tracks involve close collaboration with senior engineers and direct exposure to enterprise clients. This is not a support role; you'll be expected to contribute from day one. What You Will Learn? - Build Machine Learning models with financial data - Design, build, and maintain CRUD APIs to interact with users and serve the models - Deploy, monitor, and maintain applications in Azure and Snowflake Tech Stack - Languages: Python, SQL - ML Frameworks: PyTorch, XGBoost - API Frameworks: FastAPI - Databases: Snowflake, Postgres - Cloud: Azure, AWS How We Work? - Communication: We run one short standup every morning to align on daily priorities. Everything else lives in writing: project documentation on Linear and GitHub, async conversations on Slack with dedicated channels per team and project. - Rhythm & Organisation: We work in weekly sprints, so priorities are always clear and nothing drags. Every Friday we run a retrospective, an open conversation about what's working and what isn't. Compensation - Internships - Compensation: €830 gross/month (€5,000 gross total over 6 months) - Contract type: Convenzione Stage curriculare / extra-curriculare - Relocation bonus: €3,000 gross, paid in three instalments of €1,000 each to support your move to Milan - Food Allowance: €600 - Monthly travel reimbursement: up to €370 /month - Equipment: Brand new apple M5 14inch - High non-cash value (mentorship, fast promotion) - Duration: 6 months, with a 1-month trial period Career Path for Interns This internship is not a dead end; it's the front door. We hire interns with the explicit intention of converting them into full-time engineers. Here's what that typically looks like: - Internship → Full-Time Conversion: Most interns transition to a permanent contract within 3 to 6 months. The strongest performers make the jump in as little as 2 months. We don't believe in making people wait if the fit is clear. Compensation Full-Time - RAL: €33,000 gross, CCNL Metalmeccanico level C3 - Relocation Package: €3,000 - Food Allowance: €1,200 - Project bonuses: awarded on delivery and client impact - Year-end bonus: awarded for outstanding team performance - Salary progression: +10% at each career level - Salary reviews: it happens every year - Equipment: Brand new apple M5 14inch - Contract Type: Tempo Determinato - Duration: 6 months, with a 1-month trial period Career Path for Tempo Determinato You do a great job in your first six months with us and we will offer you a Tempo Indeterminato before the end of your contract. Remote Work & Schedule This is a remote position, and you are free to work from anywhere in Italy. However, if you fancy collaborating with other members of the team, you are welcome to join our Milan office (Via Zuretti 34, Milan). - Working hours: Monday–Friday: 8:30 – 17:30 CET - Lunch: 13:00 – 14:00 (flexible) Selection Process We keep it fast, respectful, and transparent. - Interview with CEO (15 min): Motivation, Background knowledge and Availability - Interview with CTO or Senior Engineer (15 min): Company and role presentation, alignment on expectations. - Technical Interview (30-40 min): Technical discussion on ML principles and system design. No whiteboard coding or Leetcode-style questions. Total timeline: 3 to 4 weeks Mandatory - Experience with training custom ML models using PyTorch and XGBoost - Familiarity with API development - Good understanding of relational databases and experience with querying and managing data - Knowledge of version control systems (e.g., Git) - B2+ English proficiency Nice to Have - Experience with interaction with LLMs (GPT, Claude, Gemini) via API calls - Experience with running Machine Learning inference jobs with PyTorch or ONNX Travel Once a year, the whole team flies to Dublin for a 3-day offsite at our HQ. We also invite once a year the company to a nice place.

• Responsible for the architecture of hybrid game and machine learning systems. • Design, develop, and maintain both learning and non-learning systems. • Stay up to date with the latest machine learning and game technologies. • Assist teams during development phases. • Create and maintain quality guidelines for machine learning applications. • Spread machine learning knowledge across the organization. • Improve team performance and the outcomes delivered.

• Responsible for the architecture of hybrid game and ML systems, serving as the primary contact for integrating machine learning into game systems • Design, develop and maintain learning and non-learning systems • Keep current on the latest developments in both ML and game technologies, identifying and exploring relevant applications for the company • Support programming and ML teams throughout development • Establish and maintain quality guidelines for our ML applications • Disseminate ML knowledge across the organization • Improve the function of one’s team, either through individual achievement or through leadership • Maintain rich, thoughtful, candid communication with peers in order to ensure the very best results • Gather, acknowledge, and respond to internal feedback, adjusting design and technological choices as necessary