Back

Speaker "Ali Arsanjani" Details Back

 

Topic

1. Keynote: 
When Can We Trust a Decision Made by a Machine: Building Trustable AI and Detecting Misinformation 
 
2. Workshop:
Natural Language Processing: Foundations and Applied Research

Abstract

Keynote: 
When Can We Trust a Decision Made by a Machine: Building Trustable AI and Detecting Misinformation 
 
Synopsis
With the advent of machine and deep learning, explainability and interpretability has become paramount to traceable and justifiable and explainable results:
when someone's house gets foreclosed, some one is sentenced, an insurance claim is denied, a mortgage application is denied. There are  legal, human, organizational, IP and societal implications of leveraging the augmented intelligence of machines in supporting highly complex and previously only open to highly educated, highly skilled experts in medicine, underwriting, law, adjudication, etc.
The need to create and train unbiased or minimally biased datasets for deep learning in the interests of Fairness, Accessibility and Transparency requires best practices that are not known to organizations embarking on machine learning and AI activities.
In this session we will cover the methods and techniques and best practices to detect, manage and mitigate bias in datasets, deep neural network training, application integration. We will also explore the cognitive biases that lead us to curate date to create less than trustworthy blackbox AI systems, and how to avoid them.
We will explore in detail (code level as well as architecture), the project AlternusVera which detects fakeness or deliberate misinformation in a body of text.
 
Audience:
CTOs, IT managers, Data scientists, machine learning engineers, enterprise architects
 
2. Workshop: Natural Language Processing: Foundations and Applied Research
 
In this workshop we will explore the foundations of NLP/NLU and NLG (processing, understanding and generation of natural language)
This will include a case study of Project AlternusVera which we will walk through and provide attendees with code .
AlternusVera ingests a corpus of documents and provides a misinformation index that assesses the reliability of the content against 35 factors.

 

Profile

Dr. Ali Arsanjani is the TechSector Leader and Principal Machine Learning Architect for AI/ML Solutions Architecture with Amazon Web Services . He leads the Technology Sector for Saas, Software, ISVs and Internet, where AWS largest customers are served. He also leads an Applied Science team for research in AI.
 
Previously, Ali was Founder and Vice-president of Artificial Intelligence and Machine Learning for Deep Context, an AI consultancy that provides deep contextually relevant information for customer engagement.
 
Previously, VP AI & ML at 8x8, cloud Communications. Dr. Arsanjani was responsible for research, productization and implementation of AI and ML in the UCaaS and CCaaS products.
 
(1998-2018) IBM Chief Technology Officer (CTO) for Analytics & Machine Learning, IBM Distinguished Engineer, responsible for applied research and architectural implementations leading IBM services teams in customized machine learning and analytics solutions. Building teams across multiple geos in large-scale agile solution development he was considered the father of SOA. His career spans CTO responsibilities for SOA, BPM, RPA, Analytics, Machine Learning and Artificial Intelligence Systems.
 
Ali is a hands-on machine learning executive and engineer/researcher with over 20 years experience implementing software systems that leverage service-oriented architecture, analytics and machine learning for IBM’s largest clients.
 
Ali's breadth of ML and DL expertise covers NLP/NLU/NLG, Deep learning ensemble models, anomaly/outlier/pattern detection and training, customer segmentation/churn/upsell analysis , voice/video and text analysis for conversational virtual assistant implementations.
 
Ali is also Founder of Deep Context, a deep learning startup focused on Amalgamation of data for deeper actionable insights using contextual analysis. He is an advisor to startups and boards of larger companies.
 
Ali has chaired standard bodies such as The Open Group and is responsible for co-leading the SOA Reference Architecture, SOA Maturity Model, and Cloud Computing Architecture standards.
 
In his long tenure at IBM, He & his team specialize in harvesting, developing best-practices for microservices architectures on hundreds of projects WW across multiple industries, leading a community of practice of over 6000 people.