Global Artificial Intelligence Virtual Conference 2022
2022-03-17..2022-03-19
-- online --
who it's for
machine learning engineers, data engineers, MLOps engineers, data scientists, product managers, risk and compliance officers, security engineers
topics
event-based feature engineering; time-travel feature stores; streaming graph processing; model observability and monitoring; ModelOps and deployment pipelines; federated learning and encryption; synthetic data generation; NLP embeddings and similarity search; conversational AI; privacy-preserving distributed ML; real-time adaptive machine learning; deep learning recommender systems
| title | author | length |
|---|---|---|
| Global Artificial Intelligence Virtual Conf - Webinar Mar 2022 | Anand Oka, Lakshmi Ravi, Leah McFarland, Bret Greenstein | 32m |
Multi-tiered digital twins model supply chains as networks, stateful machines and multiplayer games and enable simulated annealing optimization for inventory… | ||
| Keynote: NLP Trends and Investments | Aarthi Srinivasan | 17m |
Claims that humans send 19 billion text messages a day and generate 4 petabytes per day to justify rapid NLP demand, with the NLP market at about $12B now and… | ||
| Challenges in using machine learning for financial monitoring | Leah McFarland | 23m |
Shifting a rules-based AML system to machine learning cuts the false positive rate down ~50% and raises escalation quality, while supervised models train on… | ||
| Crouching Harm, Hidden Bias—Addressing issues of fairness.... | Ranjan Bhattacharya | 27m |
Walks through concrete failure modes: historical, representation, annotation, measurement, selection, aggregation, optimization, interpretation biases and real… | ||
| Introducing Quine: a Streaming Graph for Modern Data Pipelines | Ryan Wright | 25m |
Quine is a streaming graph interpreter that consumes Kafka/Kinesis events, builds a versioned graph per entity and stores per-node histories with append-only… | ||
| Creating & operating ML models from event-based data using feature stor... | Charna Parkey | 17m |
Claims a practical stack: a time-travel feature engine (Kaskada) that loads raw event streams like parquet files in S3 and runs in Jupyter to compute features… | ||
| Deep Recommendations for e-Commerce | Georgios Damaskinos | 23m |
Facebook and Instagram Shops launch in May 2020 and recommendations use user interaction events (clicks, likes, purchases) rather than explicit ratings, so the… | ||
| AI model management platforms | Dan Turchin | 29m |
PeopleReign is an AI‑first platform that automates the IT/HR employee service lifecycle using NLP/NLU, diagnostics, orchestration and a feedback loop that… | ||
| Cutting Edge AI Technologies of 2022 | Hina Dixit | 33m |
Gartner forecasts AI software revenue at $62.5B in 2022, up 21.3%, yet Gartner also warns ~85% of AI projects will err from bias. Concrete trends: federated… | ||
| AI-assisted Development: Providing Performance Feedback for API Programs | Malith Jayasinghe | 22m |
Claims development-time forecasts of throughput and latency for integration programs by combining per-endpoint models with a system model, using observed… | ||
| Keynote Panel : AI in Supply chain | Danny Godbout, Ajay Raghavan, Anand Oka, Ted Way | 31m |
Panel argues current supply chains fail for lack of shared, staged data across SAP/Oracle/mainframes and spreadsheets, and shows work building IoT embedded… | ||
| Managing AI Risk: Beyond Explainability | Agus Sudjianto | 28m |
Argues that high‑risk banking needs inherently interpretable models rather than black boxes with post‑hoc explainers, showing LIME and SHAP mislead on local… | ||
| Natural Language Processing (NLP) going mainstream | Alice Chung | 26m |
Claims that NLP is crossing into everyday use with no-code/low-code tools that let nonprogrammers build models, concrete deployments in social listening and… | ||
| Keynote: Building Trust in a Data Driven Enterprise | Bret Greenstein | 29m |
Argues that trust requires people-first change not just tech, recommending cloud migration, governance and literacy tied to business outcomes; cites public… | ||
| AI Observability: How To Fix Issues With Your ML Model | Danny Leybzon | 18m |
Argues that detecting drift is not enough and that logging is the key to root cause analysis, using WhyLogs profiles that compress terabytes to megabytes… | ||
| Advanced Natural Language Processing with Free Low-Code and... | David Talby | 33m |
Shows concrete tools and results: a one-line Python API (NLU) exposes over 4,500 ready models and reuses Spark NLP embeddings with 1.25 million downloads per… | ||
| Why Deep Learning is driving demand for Human in the Loop | Keith McCormick | 29m |
Explains why Imagenet 2012 and Hinton’s jump from a few hidden layers to eight catalyze deep learning’s hunger for labeled data, driving demand for human in… | ||
| Life cycle of data becoming a model | Lakshmi Ravi | 31m |
Starts from three concrete data types — behavioral, feedback, synthetic — and names the tooling and tradeoffs: build secure pipelines with Spark and Hadoop… | ||
| Process Optimization using AI/ML | Manjeet Singh | 29m |
ServiceNow demonstrates a workflow-first approach that uses process mining to map event logs in minutes (instead of months), an intent-based discovery pipeline… | ||
| Scaling privacy embedded distributed ML systems | Sharmistha Chatterjee | 31m |
Describes building a GDPR-aware pipeline for IoT and enterprise ML that combines end-to-end encryption, federated learning safeguards, and differential… | ||
| Building a Computing Environment to Support Analytics | Meghan Anzelc, Christina Fernandes D'Souza | 16m |
Spencer Stuart builds a cloud analytics environment from scratch starting in 2018, targeting internal and external structured and unstructured sources… | ||
| Time for quality ML from event-based data | Davor Bonaci | 21m |
Claims that computing features directly on raw event streams with a feature engine lets teams iterate far faster and avoid temporal leakage, using time-aware… | ||
| Employing AI to extract actionable insights from qualitative user... | Indrasis Mondal, Adam Gorski | 20m |
About 10,000 in-product comments are labeled and processed by a two-layer ensemble: a binary relevance classifier followed by a multi-class topic classifier… | ||
| Introducing ModelOps: How to deploy and manage every type of model | Marinela Profi | 26m |
Shows a concrete pipeline that retrains and replaces a production XGBoost gradient boosting model (version 4) after an alert for misclassification threshold… | ||
| Cybersecurity Data: From Linear Chains to Regulatory Compliance | Sanjay Joshi, Naveen Goela | 30m |
Claims a scalable endpoint platform that forms a linear chain peer-to-peer mesh to collect and aggregate telemetry from up to a million devices, returning… | ||
| Conversational AI | Siddharth Sureka | 25m |
Practical roadmap for conversational AI at Charles Schwab level: decide build versus buy based on strategic value, expected time to market and required NLP… | ||
| Introduction to Machine Learning Workshop | Sneha Muppalla | 01h 31m |
Hands-on intro that codes in Python using Jupyter Notebook via Anaconda, reading a CSV into pandas, encoding categorical columns with label encoding, and… | ||
| Workshop Real-time Adaptive Machine Learning Algorithms with Python | Chanchal Chatterjee | 03h 01m |
Workshop follows code from the GitHub chatterj0 repo and runs in Docker TensorFlow JupyterLab (requires ~25 GB). It implements streaming feature extraction… | ||
| Can AI make us smarter than cancer | Roushanak Rahmat | 18m |
Claim: AI improves personalized cancer care by merging genome, imaging, labs and demographics to predict treatment response, reduce trial-and-error and cut… | ||
| Can financial crashes be anticipated using physics? | Revant Nayar | 35m |
Claims that conformal field theory diagnostics flag endogenous bubble crashes by tracking a rising anomalous dimension derived from log-price correlators… | ||
| Idiomatic Pandas | Matt Harrison | 35m |
Five concrete Pandas practices produce big wins: enforce correct dtypes and convert wide integers to smaller numpy types (example converting columns to int8… | ||
| DataOps, MLOps, Computer Vision | James Le | 40m |
Argues that production computer vision fails or succeeds on data, not models, and defines three concrete data problems: curate high‑quality datapoints, label… | ||
| Data management and Analytics for Industrial IOT | Dinesh Kumar | 49m |
Explains how an on‑prem OpenTSDB + HBase on-prem stack handles high throughput (around 10,500 writes/sec in the demo and ~408,894 rows written) and contrasts… | ||
| Deep Learning-Based Human Activity Recognition for Physical Skills... | Vijay Nadkarni | 28m |
Claims an operational pipeline: IMU time series undergo statistical feature processing and clustering (kernel determinant analysis), then an LSTM classifies… | ||
| Building scalable end-to-end deep learning pipelines in the cloud | Rustem Feyzkhanov | 34m |
Argues a serverless pipeline on AWS reduces ops and idle cost by composing AWS Lambda for short parallel tasks, AWS Batch for heavy or GPU jobs, Amazon… | ||
| Keynote Panel AI from Investment perspective | bill pauling | 33m |
Panel argues venture bets shift from definitions to deployable value: robotics as a service is framed as a horizontal SaaS‑like market (investments cited in… | ||
| General Keynote Multi-tiered digital twins with what-if analysis... | Anand Oka, Mayank Shrivastava | 27m |
A Microsoft-built multi‑tier digital twin ingests systems‑of‑record, IoT telemetry and partner‑shared data into a graph and runs what‑if analysis plus… | ||
| nothing matches | ||