AWS re:Invent 2020
2020-11-30..2020-12-18
-- online --
[AWS re:Invent 2019, AWS re:Invent 2019 Global Partner Summit, AWS re:Invent 2019 | AI & Machine Learning Summit, AWS re:Invent 2020, AWS re:Invent 2021, AWS re:Invent 2022, AWS re:Invent 2023]
| title | author | length |
|---|---|---|
| Share information by removing language barriers with Amazon Translate | Watson G. Srivathsan | 31m |
Demonstrates using Amazon Transcribe + Amazon Translate to turn Open WHO videos into searchable subtitles, cutting subtitle production for a 30‑minute video… | ||
| Discovering insights from customer surveys at McDonald’s | Abhi Bhatt, Sameer Karnik | 22m |
McDonald’s pipelines millions of survey and social comments through S3 into Amazon Comprehend, uses Coreferencing to resolve pronouns and Sub-comment Breakdown… | ||
| From POC to production: Strategies for achieving machine learning at scale | Alex McClure | 34m |
Seven concrete steps for moving from POC to production: establish culture and goals, set up cross‑functional teams and a COE, secure production data with… | ||
| Get rolling with machine learning on AWS DeepRacer | Kesha Williams | 23m |
Hands-on demo of AWS DeepRacer Evo shows a six‑layer neural net trained in a 3D simulator (AWS RoboMaker) with models built in Amazon SageMaker, artifacts… | ||
| Recreating Colonel Sanders for KFC Canada using Amazon Polly Brand Voice | Robin Dautricourt | 22m |
KFC Canada gets an exclusive Colonel Sanders voice built from studio recordings of a hired impersonator and converted into a custom Amazon Polly model, then… | ||
| Getting started with generative AI with AWS DeepComposer | Rahul Suresh | 32m |
Console demo shows AWS DeepComposer converts MIDI into piano roll images and uses GANs to add accompaniment via a GAN generator vs discriminator loop, AR‑CNN… | ||
| Catch more potential online fraud faster with Amazon Fraud Detector | Nick Tostenrude | 31m |
Amazon Fraud Detector is a fully managed ML service that builds customized models from your historical CSVs (minimum 10,000 rows, ≥400 fraud examples), returns… | ||
| Get started with ML in minutes with Amazon SageMaker JumpStart | Thom Lane | 29m |
Amazon SageMaker JumpStart provides 150 pre-trained models for vision and text plus one-click fine-tune and deploy from the Studio catalog, with endpoints… | ||
| Building a successful inventory planning solution with Amazon Forecast | Namita Das, Supratim Banerjee | 35m |
Amazon Forecast claims up to 50% more accurate forecasts and adds features like a Weather Index and holiday regressors; More Retail tests DeepAR+ models and… | ||
| Deliver viewing experiences for super fans with Amazon Personalize | Mark Wood, Liam Morrison | 25m |
Demo uses MovieLens 20M (28,000 movies, 138,000 users, 20M interactions) to show Amazon Personalize trained from CSVs in S3 and served via… | ||
| Build quality ML models easily & quickly with Amazon SageMaker Autopilot | Kate Zimmerman | 32m |
Amazon SageMaker Autopilot automatically inspects tabular CSV/Redshift/Athena data, trained hundreds of models across feature engineering, algorithms (XGBoost… | ||
| How to choose the right instance type for ML inference | Shashank Prasanna | 32m |
Benchmarks show concrete tradeoffs: a ResNet50 on a G4 T4 runs TensorFlow at ~115 images/sec and ~17 ms, and after Nvidia TensorRT compilation jumps to… | ||
| Shift your ML model into overdrive with AWS DeepRacer analysis tools | Don Barber | 30m |
Announces two 2021 race divisions (open and pro) and shows concrete tuning techniques: reward-by-progress over steps (about 15 steps/sec) produces inside-line… | ||
| Using AI to automate clinical workflows | George Price | 18m |
Two ready web apps demonstrate end-to-end automation: Medical Transcription Analysis (MTA) streams audio to Amazon Transcribe Medical, runs Amazon Comprehend… | ||
| Detect machine learning (ML) model drift in production | Sireesha Muppala | 30m |
Deploys an XGBoost movie‑recommendation model trained on MovieLens, enables SageMaker data capture, generates baseline statistics.json and constraints.json… | ||
| How to use fully managed Jupyter notebooks in Amazon SageMaker | Aashmeet Kalra | 32m |
Overview of how Amazon SageMaker runs fully managed Jupyter notebooks with concrete features: one-click SageMaker Studio notebooks that spin up elastic… | ||
| Find your most expensive lines of code with Amazon CodeGuru | Daniela Tzvetkova, Behrooz Nobakht | 30m |
Amazon announces Amazon CodeGuru Reviewer and Profiler as ML-powered tools that find expensive lines of code, flag hard‑coded AWS credentials, missing DynamoDB… | ||
| How the NFL builds computer vision training datasets at scale | Vikram Madan, Sam Huddleston | 29m |
NFL video gets time-synced multi-view preprocessed, then labeled with Amazon SageMaker Ground Truth using vendor crowds plus private experts, assistive AI… | ||
| Choose the right machine learning algorithm in Amazon SageMaker | Denis V. Batalov, PhD, Alberto Danese | 30m |
Amazon SageMaker exposes 17 built-in algorithms and demonstrates an image classification pipeline using ResNet-152 with transfer learning on a public chest… | ||
| Secure and compliant machine learning for regulated industries | Stefan Natu | 29m |
Demonstrates a reproducible secure ML stack built on SageMaker Studio plus VPC endpoints, IAM condition keys, KMS keys, Service Catalog and AWS Organizations… | ||
| Intelligent document processing for the insurance industry | Raj Copparapu, Reddi Gudla | 30m |
Anthem replaces a 20‑minute manual review per claim by piping scanned claims and attachments through Amazon Textract for OCR and structure extraction, running… | ||
| Finding answers faster with intelligent search | James Rehfeld, Udi Hershkovich | 28m |
Amazon Kendra replaces an Elasticsearch keyword approach for Reg Ranger, using NLU plus multiple ML models to return direct answers, passages and ranked… | ||
| Scaling MLOps on Kubernetes with Amazon SageMaker Operators | Alex Chung, Srivathsan Canchi | 28m |
Kubernetes CRDs call out training and hosting to SageMaker Operators while Argo workflows handle GitOps CI/CD, Jenkins builds container images, and logs stream… | ||
| Train and tune ML models to the highest accuracy using Amazon SageMaker | Emily Webber | 31m |
SageMaker Studio shows concrete workflows: notebooks start in seconds, training jobs run on managed EC2 instances (including spot instances reduce cost), and… | ||
| Replicate AWS DeepRacer architecture to conquer the track with SageMaker | Sahika Genc, Ph.D. | 27m |
Replicates the AWS DeepRacer stack inside SageMaker notebooks, using RoboMaker for simulation and Kinesis Video Streams for sensor flow, and provides editable… | ||
| Fast training and near-linear scaling with DataParallel in Amazon SageMaker | Aditya Bindal | 24m |
SageMaker Distributed DataParallel removes network bottlenecks by offloading AllReduce to CPUs so the cluster’s CPUs act as CPUs-as-parameter-servers, shards… | ||
| Train ML models faster with better insights using Amazon SageMaker Debugger | Nathalie Rauschmayr, Chaim Rand | 30m |
Amazon SageMaker Debugger captures full tensors and system metrics during training, runs built‑in rules to flag overfitting, vanishing gradients, CPU/IO… | ||
| Productionizing R workloads using Amazon SageMaker, featuring Siemens | Michael Hsieh, Edgar Bahilo Rodríguez | 30m |
R code runs inside custom Docker images on SageMaker (pushed to ECR) using reticulate and the SageMaker SDK to run SageMaker processing, training, and model… | ||
| Train billion-parameter models with model parallelism on Amazon SageMaker | Can Karakus | 29m |
Shows how SageMaker Model Parallel uses MPI, automated graph/module partitioning and pipeline parallelism with micro-batches to fit and train billion-parameter… | ||
| Implementing MLOps practices with Amazon SageMaker | Shelbee Eigenbrode | 29m |
Automates a repeatable SageMaker pipeline that preprocesses US census data, trains a Scikit-learn logistic model on ML M5.xlarge, and evaluates holdout 20%… | ||
| COVID-19: Identity verification & work safety with Amazon Rekognition | Kashif Imran, Mónica Patricia Ospina Londoño | 29m |
Shows a concrete implementation using Amazon Rekognition: PPE detection API flags helmets, masks and gloves from webcam or S3 frames and returns per-person… | ||
| Detect abnormal equipment behavior by analyzing sensor data | Brant Swidler, Alberto Gascón | 28m |
Amazon Lookout for Equipment builds a custom anomaly model per asset using historical sensor data and maintenance logs and reads from S3 to train and then run… | ||
| Make sense of health data with Amazon HealthLake | Guy Eldredge | 21m |
A fully managed HIPAA-eligible service that ingests structured and unstructured records, normalizes them to FHIR v4, and appends indexed JSON with offsets and… | ||
| Amazon Lookout for Vision | Barath Balasubramanian, Shreyas Subramanian, Maikel van Verseveld | 29m |
Service automates defect detection for manufacturing using Amazon Lookout for Vision and requires as few as 30 images to bootstrap a model, builds an initial… | ||
| Enable predictive maintenance for your industrial equipment | Kartik Dalal, Magnus Akesson | 30m |
Cloud service Amazon Monitron ships purpose‑built vibration and temperature sensors plus a Wi‑Fi gateway and mobile app to start monitoring in minutes, pairing… | ||
| Use computer vision at the edge to improve operations with AWS Panorama | Mike Miller | 26m |
AWS Panorama brings real time computer vision to on‑prem cameras, claiming subsecond inference by compiling SageMaker models to run on an AWS Panorama… | ||
| How to create fully automated ML workflows with Amazon SageMaker Pipelines | Urvashi Chowdhary | 29m |
Demonstration builds a full pipeline using the SageMaker Pipelines Python SDK that runs a scikit‑learn processing step, XGBoost training job, evaluation script… | ||
| Amazon SageMaker Feature Store: Store, discover, & share features for ML apps | Vidhi Kastuar | 30m |
Amazon SageMaker Feature Store is a managed feature repository that stores features for online and offline store use, supports streaming and batch ingest… | ||
| Improve application availability w ML-powered insights using Amazon DevOps Guru | Jacob Sullivan | 27m |
Amazon DevOps Guru is an ML-powered, fully managed service that automatically pulls CloudWatch, Config, CloudTrail and X‑Ray data, auto-calibrates thresholds… | ||
| Proactively monitor health of your business using Amazon Lookout for Metrics | Ankita Verma, Brian Ecker | 28m |
Amazon releases Amazon Lookout for Metrics in preview: a fully managed ML service that ingests time series from S3 intake (also CloudWatch, Redshift… | ||
| Accelerate data preparation with Amazon SageMaker Data Wrangler | Ajai Sharma | 33m |
Cloud tool demonstrates a visual data prep workflow that connects to Athena Redshift S3 and SageMaker Feature Store and runs queries and joins without hand… | ||
| MLOps for edge devices with Amazon SageMaker Edge Manager | Vin Sharma | 32m |
SageMaker Edge Manager is discontinued April 26th, 2024 yet the session shows concrete MLOps: use SageMaker to train an autoencoder model, compile with… | ||
| Understand ML model predictions & biases with Amazon SageMaker Clarify | Michael Sun, Pinar Yilmaz | 24m |
Amazon SageMaker Clarify detects bias at three points—during data prep, after training, and continuously in production—by running a first-party processing… | ||
| Create sophisticated conversational experiences using Amazon Lex | Harshal Pimpalkhute, Caroline Territt | 30m |
Dropbox builds Dropbot with Amazon Lex, integrates it into the chat frontend and backend services, and launches it in October 2019 as an FAQ that evolves into… | ||
| Pioneering user experiences at StockX with Amazon Personalize | Blaine Dietrich, James Jory | 29m |
StockX implements Amazon Personalize with a daily ETL that writes interactions to S3, per-category dataset groups, and a retrain workflow that spins up a… | ||
| Modernize your contact center with AWS Contact Center Intelligence (CCI) | Chris Featherstone, Dave Rennyson | 31m |
AWS CCI composes Amazon Transcribe, Comprehend, Lex, Polly, Translate and Kendra into reusable building blocks while SuccessKPI wraps ~40 AWS components into a… | ||
| Intelligent document processing: Extract data & insights from documents | Sandeep Mistry | 30m |
Pipeline extracts text with Amazon Textract, runs NLP with Amazon Comprehend and routes low‑confidence cases to Amazon A2I; custom image and text classifiers… | ||
| Accelerating autonomous vehicle solutions with SageMaker, featuring Lyft | Alex Bain, Nitin Wagh | 32m |
Lyft Level 5 migrates heavy ML training from self-managed Kubernetes to Amazon SageMaker and cuts turnaround from days to minutes/hours by using distributed… | ||
| How to build personalization models using Amazon SageMaker | Chris Fregly | 33m |
Shows how to train a TensorFlow multitask recommender on MovieLens using Amazon SageMaker Studio, TensorFlow Recommenders and SageMaker hyperparameter tuning… | ||
| Build consistent and portable ML environments with containers | Antje Barth | 30m |
Prototype fine‑tuned a DistilBERT text classifier (star ratings) using Hugging Face tokenizers and TFRecords, then packages the training code in Docker, runs… | ||
| Which episode? Using AI to streamline media content operations | Anushri Mainthia, Sally Hubbard | 30m |
AWS tooling plus partner Curio turns large archives into timestamped, searchable metadata: content lands in S3, Rekognition (labels, video segment detection… | ||
| A day in the life of a machine learning data scientist at JPMorgan Chase | Daryush Laqab, Tom Lococo | 35m |
JPMorgan builds OmniAI on Amazon SageMaker as a hybrid deployment that keeps control plane and identities on‑prem while running training on AWS; the platform… | ||
| Detecting fraud faster with Amazon SageMaker, featuring Coinbase | Erik Reppel, Soji Adeshina | 27m |
Coinbase describes a production fraud stack built on Amazon SageMaker that achieves a 50% reduction in fraud year over year, uses ensemble models like CatBoost… | ||
| Modernize your call center with intelligent search using Amazon Kendra | Brian Lichtle, Sonali Sahu | 31m |
Magellan Rx consolidates documents from 13 repositories and connects Amazon Kendra to Amazon Lex, Transcribe and Comprehend to deliver real time, context aware… | ||
| Enhance speech recognition accuracy using custom language models | Paul Zhao, Javier Asensio-Cubero | 28m |
Amazon Transcribe’s custom language models train on domain text (emails, manuals, Trustpilot) uploaded to S3 and produce measurable gains — Octopus Energy… | ||
| Achieve real-time inference at scale on AWS with Deep Java Library | Zach Kimberg | 29m |
Deep Java Library (DJL) loads pretrained models from its model zoo (over 70+ models), using JNI-backed engines like PyTorch and ONNX Runtime so Java inference… | ||
| Using Amazon Transcribe to make content searchable and accessible | Esther Lee, Patrick Carroll | 29m |
Amazon Transcribe plus the Media Insights Engine turns video into searchable, timestamped text and VTT subtitles using S3, Lambda, MediaConvert, Comprehend… | ||
| Interpretability and explainability in machine learning | Vikrant Kahlir | 27m |
Focuses on using SHAP (game‑theory Shapley values) to produce local and global attributions, explains the kernel approximation that samples about 2[M]+2048… | ||
| nothing matches | ||