AI Research Engineer

Reply; Torino; In loco; EUR 55.000 - 85.000; Tempo pieno

Descrizione del lavoro

Reply is seeking an AI Research Engineer to design synthetic data pipelines, build research harnesses, and run post-training RL workflows on custom environments. The role involves distributed job scheduling across local and cloud clusters, and sharing insights via open-source platforms and model hubs.

The ideal candidate has 3–8 years of experience, strong Python, Triton and Rust skills, and a background in training LLMs with practical exposure to large-scale architectures in a hybrid or

Competenze

• Degree in Computer Engineering, AI, Math, or related field.
• Strong Python, Triton and Rust programming skills.
• 3–8 years of experience, with 2+ years training LLMs.

Mansioni

• Design synthetic dataset generation pipelines from task descriptions and production traces.
• Build auto-research harnesses to explore recipes, architectures and ablations.
• Run post-training workflows focusing on RL and agentic reasoning in custom environments.

Conoscenze

Python

Triton

Rust

LLM training

Dataset design

Formazione

Degree in Computer Engineering / AI / Math / Software Engineering

Strumenti

Hugging Face Hub

NVIDIA DGX

RL environments

Descrizione del lavoro

Are you an AI Research Engineer expert in training custom AI?

Join Reply!

WHO WE ARE

Reply specialises in the design and implementation of solutions based on new communication channels and digital media. As a network of highly specialised companies, Reply supports major industrial groups in the telecom and media; industry and services; banking and insurance and public sectors in defining and developing business models enabled by the new paradigms of AI, cloud computing, digital media and the internet of things. Reply's services include: consulting, system integration and digital services.

WHAT WILL YOU DO?

• Core activities. You will design synthetic dataset generation pipelines from task descriptions, teacher-student distillation or production traces. You will build auto-research harnesses using AI coding to navigate the design space across recipes, architectural design, dataset mixes and ablations. You will run post-training workflows, with a focus on RL and agentic reasoning, either via RL and distillation, on custom RL environments across domains for professional use on the Pareto frontier. You will schedule jobs on a distributed cluster across local compute and cloud compute and optimize serving confiturations either for online RL, evaluation or synthetic generation. You will share the core insights of your work via blog posts, open-source repositories, and on model and dataset hubs, like Hugging Face.
• Tech & Tools Stack. You will work with cloud compute clusters (Lambda, Nebius, AWS, GCP), local compute clusters (NVIDIA DGX), AI workload managers (SkyPilot), RL environments and stack (NeMo-RL, SkyRL, TRL), synthetic dataset generation pipelines (DataFlow, DataTrove), experiment trackers (MLflow), model and dataset hubs (Hugging Face Hub), evaluation metrics and environments (LM Evaluation Harness), kernels (Triton), serving (vLLM, Speculators)
• Team work. You will collaborate with a young, dynamic team of engineers and scientists in a hybrid setup. You will participate in code reviews, share knowledge, and work in a friendly, informal, and fun office environment that values creativity and continuous learning.
• Academic background. We are looking for people with a degree in Computer Engineering, AI, Math, Software Engineering, or related quantitative fields. A strong foundation in programming, especially Python, Triton and Rust, is essential, to handle training runs as well as training infrastructure.
• Technical & strategic skills. You should have from 3 to 8 years of working experience and 2+ years of experience in tracking record of training LLMs, with practical experience in either mid-training and post-training. We value candidates that have had in-depth experience on a specific training direction, such as tiny multimodal architectures for computer use, hybrid architectures, or large-scale architectures with long-context, rather than generic, surface-level notebook playtesting, and experience in crafting datasets and ablations, rather than using ready-made benchmarks and recipes.
• Nice to have. PhD in a related field and a proven track record of publication at Tier 1 conferences. Claude Code and OpenClaw are your best friends.
• Soft skills. You should be curious, proactive, and enjoy continuous learning. Strong problem-solving skills, critical thinking, adaptability, and the ability to work collaboratively in a team are important. Clear communication in English and attention to detail will help you succeed in this role.

WHAT WE OFFER

• An offer tailored to your experience. This position is open to people with varying levels of expertise and seniority. The compensation package will be determined based on your professional background, technical skills, expertise, and the level of responsibility associated with the role. The collective labour agreement (CCNL) applied is the Italian Metalworking Industry Agreement. The job classification will be assessed starting from level B2, and the gross annual salary (RAL) will range from €55,000 to €85,000. Previous experience with the training, fine-tuning, and optimization of Foundation Models, LLMs, and custom models for specific domains is considered a strong asset, along with hands‑on exposure to cutting‑edge technologies such as Agentic AI, Physical AI, and High Performance Computing.
• A structured career path. Our Career Path offers opportunities to grow both as a technical specialist and as a future leader. Your ambitions, the skills you develop, and the results you achieve will shape your professional journey.
• Continuous learning. Technology evolves rapidly - and so do we. You will join an environment that encourages curiosity, continuous learning, and the exploration of new ideas and emerging technologies.
• The benefits of being a Replyer. You will also have access to the benefits and initiatives dedicated to our community.

WHAT ARE THE NEXT STEPS

The first step of our recruiting process will be the meetings with the technical referents and then a face to face interview with the HR team. We care about an equal recruiting process.

Reply is committed to embracing diversity and creating an inclusive work environment by valuing the uniqueness of people regardless of age, gender, sexual orientation, religion, nationality, or disabilities as protected by Italian Law (L.68/99).

Furthermore, Reply is committed to ensuring a fair and accessible selection process: to help you during the recruitment process, please let us know of any kind of support you may need.