Shashank Prasanna
Developer Relations Leader | AI/ML Technical Expert | Developer Marketing Strategist | Technology EducatorDocendo discimus (by teaching, we learn)
āāāāāāāāāāāāāāāāā-Seneca
Hi!šš½ Iām Shashank ā I build, grow, and inspire AI/ML developer communities around the world.
With over 15 years leading developer relations and marketing programs at Apple, NVIDIA, AWS, Meta, Modular AI and MathWorks, Iāve helped thousands of developers adopt new AI/ML frameworks, AI cloud services and infrastructure, on-device AI frameworks, and AI accelerators for training and inference.
Iām a hands-on educator known for delivering deeply technical talks, running hands-on workshops, consulting developers on AI/ML best practices, writing clear developer messaging, and launching developer-first products that make advanced AI/ML frameworks, libraries and tools more accessible. From planning global developer engagement strategies to coaching speakers and producing content for major events such as WWDC, GTC and re:Invent my focus is simple: help developers succeed, and build communities that last.
I currently work at Apple as an AI/ML Technologies Evangelist. Here, I plan and lead the AI/ML developer content at WWDC, shape messaging for new frameworks and APIs, and coach speakers to tell their stories with clarity and impact. I also educate app developers on how to create groundbreaking AI/ML-driven apps on Apple platforms.
I believe the best developer programs are built on three things: (1) the highest-quality technical education, (2) a relentless focus on developer productivity, and (3) consistent and honest community connection and sharing.
When Iām not writing, reading, or sharing what I learn, youāll find me running š local streets and trails, testing the latest running shoes, reading running books, or brewing the perfect cup of coffee āļø.
Featured Talks, Publications and Videos
Iāve spoken at events, conferences and meetups across the globe, including NVIDIA GTC
AWS re:Invent
Apple Developer Center Events
Open Data Science Conference (ODSC)
Docker events
Future Technologies Conference (FTC)
O’Reilly AI
AWS re:MARS
AWS Summits
AWS Innovate
Dev Days
Collision Conference
AI Accelerator Conf.
University & Community Meetups
and at numerous community meetups and university events.
Youāll find many of those talks, workshops, YouTube videos and published articles below.
 Choosing the right GPU for deep learning on AWS (š„150k+ views)
 A complete guide to AI accelerators for deep learning inference (š„50k+ views)
 AI accelerators, machine learning algorithms and their co-design and evolution
 How Docker Runs Machine Learning on NVIDIA GPUs, AWS Inferentia, and Other Hardware AI Accelerators
 Deploying GPU-Optimized Machine Learning Models to the Cloud and the Edge
 TensorRT 3: Faster TensorFlow Inference and Volta Support
 Production DeepĀ LearningĀ with NVIDIA GPU Inference Engine
 DeepĀ LearningĀ for Computer Vision with MATLAB and cuDNN
 DeepĀ LearningĀ for Object Detection with DIGITS
 GTC: A Developerās Guide to Choosing the Right GPUs for Deep Learning
 GTC: A Developerās Guide to Improving GPU Utilization and Reducing Deep Learning Costs
 GTC: Improve ML Training Performance with Amazon SageMaker Debugger
 Accelerating Data Science with NVIDIA RAPIDS on AWS
 GTC: GPU-Accelerated Deep Learning at Scale with TensorFlow, PyTorch, and MXNet in the Cloud
 Why use Docker containers for machine learning development?
 Introducing Amazon SageMaker Components for Kubeflow Pipelines
 A quick guide to distributed training with TensorFlow and Horovod
 Choose the right instance for inference deployment with SageMaker Inference Recommender
 Speeding up deep learning training with SageMaker Training Compiler
 Introduction to Amazon SageMaker Serverless Inference | Concepts & Code examples
 Get access to FREE GPU-powered JupyterLab based IDE for ML with Amazon SageMaker Studio Lab
 Machine Learning with Containers and Amazon SageMaker
 Keynote: How machine learning is making customer experience more human
 Using Containers for Deep Learning Workflows
 Train Deep Learning Models on GPUs using Amazon EC2 Spot Instances (outdated approach)
 7 things you should know about AI & Machine Learning launches at re:invent 2021
 How Pytorch 2.0 accelerates deep learning with operator fusion and CPU/GPU code-generation
 Machine learning with AutoGluon, an open source AutoML library
 Introduction to TorchServe, an open-source model serving library for PyTorch
 Deploying PyTorch models for inference at scale using TorchServe
 A quick guide to managing machine learning experiments
 How to scale machine learning experiments
 How to debug machine learning models to catch issues early and often
 An easy introduction to Mojoš„ for Python programmers
 Implementing NumPy style matrix slicing in Mojoš„
 How to setup a Mojoš„ development environment with Docker containers
 Introduction to Tensors in Mojoš„
 Using the Mojo š„ Visual Studio Extension š
 Using Mojoš„ with Docker containers
 Getting started with the Mojo SDKš„
 Speeding up Python code with Mojoš„: Mandelbrot example
 Whatās New in Mojo 24.4: Improved Collections, New Traits, OS Module Features, and Core Language Enhancements
 Fast K-Means Clustering in Mojo: Guide to Porting Python to Mojo for Accelerated K-Means Clustering
 Whatās New in Mojo 24.3: Community Contributions, Pythonic Collections, and Core Language Enhancements
 Row-Major vs Column-Major Matrices: A Performance Analysis in Mojo and NumPy
 Whatās New in Mojo 24.2: Mojo Nightly, Enhanced Python Interop, OSS Stdlib, and More
 Deploying MAX on Amazon SageMaker
 Mojo Pi: Approximating Pi with Mojo Using Monte Carlo Methods
 Evaluating MAX Engine Inference Accuracy on the ImageNet Dataset
 Optimize and Deploy AI Models with MAX Engine and MAX Serving
 Getting Started with MAX Developer Edition
 What Are Dunder Methods? A Guide in Mojo
 Mojo & Python: Calculating and Plotting a Valentineās Day Using Mojo and Python
 What Is Loop Unrolling? How You Can Speed Up Mojo
 Mojo SDK v0.7 Now Available for Download
š„ Mojoš„ livestreams
 Modular Community Livestream - New in MAX 24.4
 Modular Community Livestream - New in MAX 24.3
 Modular Community Livestream - New in MAX 24.2
 Modular Community Livestream - MAXā”ļøDeveloper Edition!
 Modular Community Livestream - Mojoš„ SDK v0.7 edition!
 Modular Community Livestream – ModCon recap + Q&A
 Modular Community Livestream - Mojoš„ SDK v0.5 edition!
 Modular Community Livestream - Mojoš„ on Mac
 Modular Community Livestream - Mojoš„ SDK
 Modular Community Q&A Livestream
š„ PyTorch livestreams
 PyTorch 2.0 Live Q&A Series: PyTorch 2.0 Export
 PyTorch 2.0 Live Q&A Series: A Deep Dive on TorchDynamo
 PyTorch 2.0 Q&A: Deep Dive into TorchInductor and PT2 Backend Integration
 PyTorch 2.0 Q&A: Optimizing Transformers for Inference
 PyTorch 2.0 Q&A: Dynamic Shapes and Calculating Maximum Batch Size
 PyTorch 2.0 Q&A: TorchRL
 2-D Parallelism using DistributedTensor and PyTorch DistributedTensor
š ļø Workshops
 A Tour of PyTorch 2.0
 PyTorch Distributed Training on AWS
š£ Announcements & Company News
 Key announcements from ModCon 2023
 What’s new in Mojo SDK v0.5?
 Mojoš„ is now available on Mac
 What’s the difference between the AI Engine and Mojo?
 Modular to bring NVIDIA Accelerated Computing to the MAX Platform
 Modular partners with Amazon Web Services (AWS) to bring MAX Enterprise Edition exclusively to AWS services
 ModCon 2023 sessions you don’t want to miss!



