Bruno Pistone

Bruno Pistone

WW GenAI Lead LLM Customization | SageMaker AI @ AWS

Proven leader in foundation model customization, distributed training, and agentic AI platform development.

Named among the Top 500 Italians Shaping the Future of AI (2024)

Impact

100+

Enterprise customers transformed with GenAI & LLM solutions

30+

Technical publications on ML & distributed systems

2M+

Developers using open-sourced training frameworks

50+

AI specialists in AWS technical community led

Expertise

Deep Learning

PyTorch Hugging Face DeepSpeed Megatron-LM FSDP LoRA/QLoRA PEFT

Distributed Systems

torch.distributed NCCL Ray MPI SageMaker HyperPod Kubernetes (EKS)

Optimization

Mixed Precision (BF16, FP16, FP8) Gradient Checkpointing Activation Checkpointing ZeRO

Accelerators & Infra

NVIDIA GPUs AWS Trainium AWS Inferentia Multi-node GPU Clusters

GenAI Frameworks

LangChain LlamaIndex CrewAI Semantic Caching

Search & Storage

FAISS pgvector Amazon OpenSearch

Evaluation & Safety

RAGAS LLM Evaluation Pipelines Prompt Injection Testing Guardrails

MLOps & Production

CI/CD for ML Model Monitoring Inference at Scale SageMaker Bedrock

Experience

WW GenAI Lead LLM Customization – Amazon SageMaker AI

Amazon Web Services

2024 – Present

Worldwide Solutions Architect leading LLM customization on Amazon SageMaker AI. I own the technical go-to-market strategy for model customization, fine-tuning, optimized inference, and agentic AI - and build the feature prototypes that engineering productizes into the service, including Serverless model customization and Ray as a distributed training option on SageMaker Training jobs. Founded and lead a global field community of 50+ AI specialists, and shape the product roadmap across 10+ major feature launches with product and engineering teams. Pioneered distributed training for 70B+ parameter LLMs — cutting training time by 30% and memory footprint by 40% — and authored 30+ technical blogs and open-source solutions now used as reference architectures for RAG, agentic AI, and foundation model training.

Sr. Generative AI/ML Specialist SA – Global Strategic Accounts

Amazon Web Services

2023 – 2024

Drove GTM strategy for GenAI across Automotive, Manufacturing, and Financial Services. Designed RAG systems processing 1M+ documents with sub-200ms latency.

ML Engineer → AI/ML Specialist SA

Amazon Web Services

2020 – 2023

Built production ML platforms, established MLOps standards across EMEA, launched multi-tenant ML platforms for 100+ data scientists.

Senior Data, ML & Cloud Engineer

Machine Learning Reply

2019 – 2020

Developed production ML models on GCP, led computer vision R&D, mentored engineering teams.

More on the Resume →

Featured Publications

Talks & Speaking

AWS re:Invent — 4× speaker

AWS re:Invent

Las Vegas, USA

• Customize small language models for specific domains
• Streamline the AI model development lifecycle with Amazon SageMaker AI
• Mastering foundation model customization with Amazon SageMaker AI
• Customize Amazon Nova models for agentic AI with SageMaker AI and Amazon Bedrock

AWS Summits — London, Milan & New York

AWS Summit

2021 – Present

• Improve the performance, quality, and availability of ML applications through AI
• GenAI for the prevention of workplace risks (with a manufacturing enterprise)
• Fine-tune LLMs with Amazon SageMaker HyperPod
• Customize Amazon Nova on Amazon SageMaker AI for your business needs

AI conferences — AI Week Italy & NLP Conference

AI Week Italy 2025 · NLP Conference 2023, London

Italy & UK

• Amazon SageMaker AI: build, train, deploy, and manage GenAI & ML models at scale — AI Week Italy
• Serverless Retrieval Augmented Generation — NLP Conference

Enterprise keynotes & partner sessions

Automotive, manufacturing & financial services · partners (Hugging Face, NVIDIA)

EMEA · 2022 – Present

• Generative AI on AWS: sharing the future of innovation — with Hugging Face & NVIDIA
• Invited keynotes for leading automotive and manufacturing enterprises

Contact