View
4
Download
0
Category
Preview:
Citation preview
Closing SessionAyan BhattacharyaCognitive Computing Thought LeaderSep 23 2016
2
A decade of being involved in Legal Matters, ranging from providing access to basic legal information to the world’s largest democracy to using AI for next gen Legal
Speaker Profile – Global Technologist
2006Completed MBA in
Washington DC
2016Cognitive Leader
Deloitte Consulting
3
The intersection of Legal practitioners, AI capabilities and society’s adoptionLegal protection and technology innovation
Artificial Intelligence – How do we define it? Where are we seeing investments?
Why is the Legal domain embracing advanced analytics?
What are some of the societal expectations and technology adoption?
4
Large law firms in the US have been dealing with consolidation in an environment where legal tech startups thrive, and general counsels are taking on more activities
Industry case loads for Transactional and Litigation workload
Source: Thomson Reuters, 2016 Data
5
Needs are stabilizing, but there is a growing trend of clients negotiating Billable times The Legal market is returning to the New Normal
Growth in Demand for Law Firm Services Billed and Collected
6
Consumers are using their historic legal service needs to drive selection of the appropriate engagement models with a trend towards fixed fee agreements
Legal service providers are embracing different commercial options
Value-based Retainer Project Unit-based
Client retains services based on individual or project requirements. Each project is billed separately by the Legal Service provider
Client retains services and pays per unit of work or deliverable
Client retains dedicated service hours per month to manage campaign requests. Additional hours are billed as incremental
Client and Legal providers jointly define value-based metrics that define payment structure. Both companies must agree to the metrics
Examples:Flat-fee per new lead acquired% of revenue increase
Examples:Fixed monthly cost to retain a campaign delivery team
Examples:Campaign requirements are estimated and billed as a one-time project
Examples:Fixed fee per rate schedule of services
Models:Variable based on metrics
Models:Fixed feeTime and materials
Models:Fixed feeTime and materials
Models:Fixed fee
7
Productivity solutions are being embraced by the Legal profession in personal and professional areas and resulting in speedier drafting and search functions
Legal Innovation is happening across the board globally in a few different ways
Legal Managed Services
Info Providers and Workflow Solutions
Thomson Reuters Westlaw
Lexis Nexis
Law firms investing in Legal Tech Startups
8
The intersection of Legal practitioners, AI capabilities and society’s adoptionLegal protection and technology innovation
Artificial Intelligence – How do we define it? Where are we seeing investments?
Why is the Legal domain embracing advanced analytics?
What are some of the societal expectations and technology adoption?
9
A quick overview on the key areas for technology exploration in Legal TechnologyAI - Unraveled
• Handle massive amounts of consumer purchase data to derive insights
• Evaluate customer preferences across channels
Big Data
• Build compelling geographic visuals with disparate data sources
• Knowledge graphs to visualize relationships between diverse data sets
Advanced Visualization• Improve analytics to
allow legal professionals to mine historical datasets
• Implement strategic data roadmap to prepare data for enhanced search
Core Modernization
Cognitive advantage
Advanced Data MgtCore Capabilities
• Develop applications to respond to customer requests with natural language
• Extract insights from semi and unstructured data from internal and external sources
10
The Legal domain like many industries is looking to develop ontologies for domain adaptation that can ultimately be used to train machines to infer and reason
Unstructured data and Natural Language Classification
• Customer service rep notes (e.g., call center)
• Emails
• Consumer complaint data
• Call center transcript
• Online forum discussions
• Email exchanges
• Online chat support
• Medical documents
• Legal filings
• Audio
• Video
Data Sources
Conversations Transcriptions
• Blogs
• News articles
• Chat transcripts
Open text External content
Analytics Capabilities
ExtractExtract information from text samples and narratives through classification in an automated fashion
Process narratives via sentence breaking/ segmentation, PoS tagging, word ambiguity resolution
Explore and visualize narratives, sentences and lexical relationships (synonyms, antonyms)
Analysis of relationships and similarity between text samples
Process
Explore
Analyze
11
• Supervised learning: the machine is presented with example inputs and their desired target outputs and the goal is to learn a general rule that maps inputs to outputs
• Regression problems if outputs are continuous
• Classification problems if outputs are discrete
• Unsupervised learning: no target outputs are given to the machine, leaving it to discover hidden patterns in data. A typical unsupervised learning problem is clustering, i.e. discovering groups of data points that are in some way similar
• Reinforcement learning: the machine interacts with a dynamic environment to maximize a numerical reward signal without a teacher explicitly telling it whether it has come close to its goal. Instead, the machine must discover which actions yield the most reward by trying them
We are seeing innovation in 3 areasMachine Learning
Exponential Growth of Data
Smarter Algorithms
Faster Processing Speed
12
Data Scientists are discovering new insights and patterns from historical data setsData Mining
RegressionAttempting to find a predictive function which models the data with the least error
ClassificationTask of generalizing known structure to apply to new data, e.g. distinguishing legitimate Vs spam emails Summarization
More compact representation of the data set including visualization and report generation
Anomaly DetectionIdentification of unusual data records, that might be interesting and require further investigation
ClusteringIn Data Mining domain task of discovering groups and structures that are similar
Association rule learningDependency modeling –searches for relationships between variables
13
These range from machine learning to virtual advisors / botsThere are over 2000 companies which are operating in Augmented Intelligence
14
Developing Industry solutions using Advanced AnalyticsDeloitte’s Cognitive Advantage Perspective
KnowYour Customer
Automate Complex Processes
Cognitive Insights
Cognitive Automation
Enable a machine to replicate human actions and judgement by leveraging robotics and cognitive technologies
Cognitive Engagement
Provide contextual and timely insights to end users using intelligent agents to drive desired actions and outcomes
Enable employees to improve the quality, accuracy, and timeliness of service by seamlessly scanning large amounts of data for answers
Build deep understanding of customers through segmentation, customer 360 profiles, personalized insights,
Improve health outcomes and costs for patients by integrating real-time data to monitor and engage stakeholders at moments that matter
Discover the next big trend
Amplify organizational intelligence by delivering deep, actionable, and real-time insights through pattern detection and analysis of billions of data sources
Make Informed Decisions
15
The intersection of Legal practitioners, AI capabilities and society’s adoptionLegal protection and technology innovation
Artificial Intelligence – How do we define it? Where are we seeing investments?
Why is the Legal domain embracing advanced analytics?
What are some of the societal expectations and technology adoption?
16
Australia has a very comprehensive program spanning Legal to all government agencies; Singapore’s Tax Authority has a virtual advisor answering questions
Governments are utilizing technology in varying levels to provide access to the masses
17
Individuals trust technology more than humans, and are more trusting with their needsMillennials are accustomed to a Trust economy and rely on technology rather than govt
18
Recommended