Artificial Intelligence Strategy in 2019 - From Experimentation to Implementation

Artificial intelligence has moved beyond the hype cycle to become a transformative business technology. Organizations across industries are transitioning from isolated AI experiments to strategic implementation, embedding intelligent capabilities into core business processes and customer experiences. This comprehensive guide explores the current state of AI in business, highlighting practical applications, implementation approaches, and how organizations can develop effective AI strategies that deliver measurable value.

The Current State of AI in Business

Where the technology stands today:

• Mainstream Adoption: 37% of organizations have implemented AI in some form
• Practical Focus: Shift from speculative use cases to proven applications
• Operational Integration: Moving from standalone projects to process embedding
• Democratized Access: Expanding AI capabilities beyond data scientists
• Measurable Results: Growing evidence of business impact and ROI

These developments are shaping how organizations approach AI implementation in 2019 and beyond.

AI Technology Fundamentals

Key concepts for business leaders:

1. Machine Learning

Algorithms that learn from data:

• Supervised Learning: Training on labeled examples to make predictions
• Unsupervised Learning: Finding patterns in unlabeled data
• Reinforcement Learning: Learning through trial and error with feedback
• Deep Learning: Neural networks with multiple layers for complex patterns
• Transfer Learning: Applying knowledge from one domain to another

Business Impact: Machine learning enables organizations to automate decisions, predict outcomes, and discover insights that would be impossible with traditional analytics.

2. Natural Language Processing

Understanding and generating human language:

NLP Capabilities

• Text Classification: Categorizing documents and messages
• Sentiment Analysis: Determining emotional tone in text
• Named Entity Recognition: Identifying people, organizations, and locations
• Language Translation: Converting between different languages
• Text Generation: Creating human-like written content

Business Impact: NLP allows organizations to analyze unstructured text data, automate communication, and create more natural human-computer interactions.

3. Computer Vision

Extracting meaning from images and video:

Vision Capabilities

• Image Classification: Identifying objects and scenes
• Object Detection: Locating and labeling multiple items
• Facial Recognition: Identifying individuals from facial features
• Optical Character Recognition: Converting images of text to editable text
• Video Analysis: Understanding activities and events in video

Business Impact: Computer vision enables automation of visual inspection tasks, enhances security systems, and creates new ways for customers to interact with products and services.

4. AI Infrastructure and Platforms

Technology supporting AI development:

Infrastructure Elements

• Cloud AI Services: Managed machine learning platforms
• AI Development Frameworks: Tools for building custom models
• Data Processing Pipelines: Systems for preparing and managing data
• Model Management: Versioning, deployment, and monitoring
• Edge AI: Running models on local devices rather than in the cloud

Business Impact: Modern AI infrastructure reduces the technical barriers to implementation, enabling organizations to develop and deploy AI solutions more quickly and cost-effectively.

High-Value Business Applications

Practical AI use cases delivering measurable value:

1. Customer Experience Enhancement

Personalizing and improving interactions:

Key Applications

• Intelligent Recommendations: Suggesting relevant products and content
• Conversational Interfaces: Chatbots and virtual assistants
• Customer Insights: Understanding preferences and behavior patterns
• Personalized Marketing: Tailoring messages to individual customers
• Sentiment Monitoring: Tracking customer attitudes across channels

Implementation Example: Sephora's Visual Artist uses AI-powered computer vision to allow customers to virtually try on makeup products, increasing engagement and conversion rates by 30% for featured items.

2. Operational Efficiency

Automating and optimizing processes:

Key Applications

• Predictive Maintenance: Anticipating equipment failures
• Intelligent Document Processing: Extracting information from unstructured documents
• Supply Chain Optimization: Improving forecasting and logistics
• Quality Control Automation: Detecting defects and anomalies
• Resource Allocation: Optimizing staffing and asset utilization

Implementation Example: Siemens has implemented AI-based predictive maintenance across its gas turbine fleet, reducing unplanned downtime by 30% and maintenance costs by 20%.

3. Decision Support and Augmentation

Enhancing human decision-making:

Key Applications

• Risk Assessment: Identifying potential threats and opportunities
• Fraud Detection: Spotting suspicious patterns in transactions
• Market Intelligence: Analyzing competitive landscape and trends
• Clinical Decision Support: Assisting healthcare diagnostics
• Financial Analysis: Uncovering insights in complex financial data

Implementation Example: JPMorgan Chase's COIN (Contract Intelligence) platform uses NLP to review legal documents, completing in seconds work that previously took 360,000 hours of lawyer time annually.

4. Product and Service Innovation

Creating new offerings and capabilities:

Key Applications

• AI-Enhanced Products: Adding intelligent features to existing offerings
• Predictive Services: Anticipating customer needs before they arise
• Automated Content Creation: Generating personalized media and communications
• Smart Environments: Spaces that adapt to user behavior
• Data-as-a-Service: Insights and intelligence as marketable products

Implementation Example: John Deere has developed AI-powered precision agriculture tools that analyze data from sensors, drones, and satellites to provide farmers with field-level insights, increasing crop yields by up to 15% while reducing fertilizer and pesticide use.

Industry-Specific AI Applications

How different sectors are leveraging AI:

1. Financial Services

Transforming banking and investment:

• Algorithmic Trading: Automated market analysis and execution
• Credit Scoring: More accurate assessment of creditworthiness
• Anti-Money Laundering: Detecting suspicious transaction patterns
• Personalized Banking: Tailored financial advice and services
• Claims Processing: Automating insurance claims assessment

Example: HSBC has implemented AI-based anti-money laundering systems that have reduced false positives by 20% while increasing the detection of genuinely suspicious activity, significantly improving compliance efficiency.

2. Healthcare and Life Sciences

Enhancing patient care and research:

• Diagnostic Assistance: Supporting clinical decision-making
• Drug Discovery: Accelerating pharmaceutical research
• Patient Monitoring: Tracking health status and predicting issues
• Treatment Personalization: Tailoring therapies to individual patients
• Administrative Automation: Streamlining healthcare operations

Example: Mayo Clinic is using AI to analyze ECG readings, identifying previously undetectable heart conditions with 87% accuracy, enabling earlier intervention and improved patient outcomes.

3. Manufacturing and Industrial

Improving production and operations:

• Smart Factories: Autonomous and adaptive production systems
• Product Design Optimization: Generative design for improved performance
• Quality Assurance: Automated visual inspection and testing
• Energy Optimization: Reducing consumption in industrial processes
• Inventory Management: Predictive stocking and distribution

Example: BMW has implemented AI-powered quality control systems in its manufacturing plants, using computer vision to detect defects with 99% accuracy, reducing inspection costs while improving product quality.

4. Retail and Consumer Goods

Enhancing shopping experiences:

• Demand Forecasting: Predicting product demand patterns
• Visual Search: Finding products from images
• Dynamic Pricing: Optimizing prices based on multiple factors
• Inventory Optimization: Ensuring product availability
• Customer Journey Analysis: Understanding shopping behavior

Example: Stitch Fix combines AI algorithms with human stylists to provide personalized clothing recommendations, analyzing over 100 attributes per item and customer preferences to create curated selections.

Developing an AI Strategy

Approaches for effective AI implementation:

1. Business Opportunity Identification

Finding high-value applications:

Identification Approaches

• Pain Point Analysis: Addressing significant business challenges
• Value Chain Assessment: Examining each step of core processes
• Customer Journey Mapping: Identifying friction points in experiences
• Competitive Intelligence: Understanding industry AI adoption
• Data Asset Evaluation: Leveraging existing information resources

Best Practice: Create a prioritized portfolio of AI opportunities based on potential business impact, implementation feasibility, and strategic alignment.

2. Data Strategy Development

Building the foundation for AI:

Data Strategy Elements

• Data Governance: Establishing policies and responsibilities
• Data Architecture: Designing systems for AI readiness
• Data Quality Management: Ensuring accurate, complete information
• Data Integration: Connecting disparate information sources
• Data Ethics: Responsible collection and use practices

Best Practice: Develop a comprehensive data strategy that treats information as a strategic asset, with clear governance, quality standards, and integration approaches.

3. Organizational Capability Building

Developing AI competencies:

Capability Elements

• Talent Strategy: Acquiring and developing AI skills
• Operating Model: Structuring teams and responsibilities
• Technology Infrastructure: Building technical foundations
• Process Integration: Embedding AI into workflows
• Change Management: Preparing the organization for transformation

Best Practice: Create a balanced approach to capability building that combines hiring specialized talent, upskilling existing employees, and strategic partnerships with external experts.

4. Implementation Approach

Executing AI initiatives:

Implementation Models

• Center of Excellence: Centralized AI expertise and governance
• Embedded Teams: AI specialists within business units
• Hybrid Model: Combining centralized and distributed capabilities
• Partnership Ecosystem: Collaborating with external providers
• Acquisition Strategy: Buying AI capabilities and talent

Best Practice: Adopt an agile, iterative implementation approach that delivers incremental value while building toward more sophisticated capabilities over time.

AI Technology Ecosystem

The developing AI landscape:

1. AI Platform Providers

Core AI infrastructure and services:

Key Players

• Google Cloud AI: Machine learning and specialized AI services
• Amazon Web Services: Comprehensive AI and ML capabilities
• Microsoft Azure AI: Enterprise-focused AI platform
• IBM Watson: Business-oriented AI solutions
• Specialized Providers: Vertical-specific AI platforms

Strategic Approach: These companies are making AI more accessible through cloud-based services, pre-trained models, and development tools that reduce implementation complexity.

2. AI Software and Solutions

Specialized AI applications:

Solution Categories

• Conversational AI: Chatbots and virtual assistants
• Computer Vision Systems: Image and video analysis
• Predictive Analytics: Forecasting and optimization
• Process Automation: Intelligent workflow tools
• Industry-Specific Solutions: Vertical applications

Strategic Approach: AI software providers are creating increasingly sophisticated solutions that address specific business problems, often requiring minimal technical expertise to implement.

3. AI Research and Innovation

Advancing the state of the art:

Innovation Sources

• Academic Institutions: Universities and research centers
• Corporate Research Labs: Company-funded AI research
• Open Source Communities: Collaborative development
• Government Initiatives: Public sector AI programs
• Startups and Incubators: Entrepreneurial innovation

Strategic Approach: Organizations are increasingly collaborating with research institutions and open source communities to access cutting-edge AI capabilities and talent.

4. AI Ethics and Governance

Responsible AI development:

Governance Elements

• Ethics Frameworks: Principles for responsible AI
• Bias Mitigation: Addressing algorithmic fairness
• Transparency Tools: Explaining AI decisions
• Privacy Protection: Safeguarding personal information
• Regulatory Compliance: Meeting legal requirements

Strategic Approach: Leading organizations are developing comprehensive AI governance frameworks that ensure ethical, transparent, and accountable use of the technology.

Overcoming AI Implementation Challenges

Addressing common obstacles:

1. Data Quality and Availability

Managing information challenges:

• Challenge: Insufficient or poor-quality data for AI models
• Solutions:
  • Implement systematic data quality processes
  • Create synthetic data for training when necessary
  • Leverage transfer learning to reduce data requirements
  • Develop incremental data collection strategies
  • Establish clear data governance and standards

Example: American Express developed a comprehensive data quality framework for its AI initiatives, establishing automated monitoring and remediation processes that improved model performance by 15% through better training data.

2. Integration with Legacy Systems

Connecting AI with existing infrastructure:

• Challenge: Difficulty incorporating AI into established technology
• Solutions:
  • Implement API-based integration approaches
  • Create middleware layers between AI and legacy systems
  • Develop phased modernization strategies
  • Use robotic process automation as a bridge
  • Leverage cloud services for flexible integration

Example: Anthem Health Insurance created an integration layer that connects its AI-powered claims processing system with legacy databases, enabling intelligent automation without replacing core systems.

3. Talent and Skills Gaps

Building necessary capabilities:

• Challenge: Shortage of AI expertise and experience
• Solutions:
  • Develop internal training and upskilling programs
  • Create partnerships with universities and research institutions
  • Leverage managed AI services to reduce technical requirements
  • Implement collaborative human-AI workflows
  • Establish AI centers of excellence to share knowledge

Example: JPMorgan Chase created an AI training program that has upskilled over 300 existing employees in machine learning and data science, building internal capabilities while addressing the talent shortage.

4. Change Management and Adoption

Ensuring organizational acceptance:

• Challenge: Resistance to AI-driven changes in work processes
• Solutions:
  • Focus on augmenting rather than replacing human workers
  • Demonstrate clear benefits to end users
  • Involve stakeholders in the development process
  • Provide comprehensive training and support
  • Celebrate and publicize early successes

Example: UPS involved its drivers in the development of its ORION route optimization system, incorporating their knowledge into the AI and creating champions who helped drive adoption across the organization.

Measuring AI Success

Approaches to evaluating effectiveness:

1. Business Impact Metrics

Assessing commercial outcomes:

• Revenue Growth: Increased sales from AI-enhanced offerings
• Cost Reduction: Operational savings from automation and optimization
• Productivity Improvement: Enhanced output per employee
• Customer Retention: Reduced churn through better experiences
• Market Share Gains: Competitive advantage from AI capabilities

Best Practice: Establish clear baseline measurements before implementation and track improvements over time, focusing on metrics most relevant to your strategic objectives.

2. Operational Metrics

Measuring process improvements:

• Process Cycle Time: Reduction in end-to-end duration
• Error Reduction: Decrease in mistakes and exceptions
• Resource Utilization: Improved use of assets and personnel
• Decision Quality: Better outcomes from AI-augmented decisions
• Response Time: Faster reaction to events and opportunities

Best Practice: Identify specific operational KPIs that AI should influence and measure both before and after implementation to quantify the direct impact.

3. AI-Specific Metrics

Evaluating technical performance:

• Model Accuracy: Correctness of AI predictions and classifications
• Precision and Recall: Balance between false positives and negatives
• Inference Time: Speed of AI decision-making
• Bias Metrics: Fairness across different groups
• Explainability Measures: Transparency of AI reasoning

Best Practice: Develop a balanced scorecard of technical metrics that ensure AI systems are not only accurate but also fair, explainable, and performant in production environments.

4. Innovation Metrics

Assessing transformative impact:

• New Product Development: AI-enabled offerings created
• Process Reinvention: Workflows fundamentally transformed
• Business Model Innovation: New ways of creating and capturing value
• Intellectual Property Generation: Patents and proprietary AI assets
• Ecosystem Development: Partnerships and platforms established

Best Practice: Look beyond incremental improvements to measure how AI is enabling fundamental innovation and transformation in your business model and operations.

The Future of AI in Business: 2019 and Beyond

Emerging trends and developments:

1. AI Democratization

Making AI accessible to more users:

• AutoML: Automated machine learning development
• No-Code AI Platforms: Visual tools for AI creation
• Pre-Trained Models: Ready-to-use AI capabilities
• AI Marketplaces: Exchanging models and components
• Embedded AI: Intelligence built into standard software

Strategic Implication: AI will become a standard capability available to all knowledge workers rather than a specialized technology requiring rare expertise.

2. AI and Emerging Technologies

Convergence with other innovations:

• AI + IoT: Intelligent processing of sensor data
• AI + Blockchain: Trusted and transparent intelligent systems
• AI + 5G: Real-time distributed intelligence
• AI + Edge Computing: Local processing for immediate decisions
• AI + Quantum Computing: Solving previously impossible problems

Strategic Implication: The most powerful applications will come from combining AI with other emerging technologies, creating capabilities greater than any single innovation.

3. Human-AI Collaboration

Evolving relationship between people and machines:

• Augmented Intelligence: AI enhancing human capabilities
• Explainable AI: Systems that can justify their recommendations
• Adaptive Interfaces: Experiences that adjust to individual users
• Collaborative Workflows: Seamless handoffs between humans and AI
• Continuous Learning Systems: AI that improves through human feedback

Strategic Implication: The future of work will involve increasingly sophisticated collaboration between humans and AI, with each focusing on their complementary strengths.

4. Responsible AI Evolution

Ethical and sustainable development:

• Algorithmic Fairness: Ensuring equitable AI outcomes
• Privacy-Preserving AI: Learning without compromising personal data
• Sustainable AI: Reducing the environmental impact of computation
• Value Alignment: Ensuring AI systems reflect human values
• Global AI Governance: International frameworks for responsible use

Strategic Implication: Organizations will need to develop comprehensive approaches to responsible AI that address ethics, fairness, transparency, and sustainability.

Conclusion: Strategic Imperatives for Business Leaders

As we progress through 2019, artificial intelligence is transitioning from experimental technology to essential business capability. While challenges remain, organizations across industries are implementing AI solutions that deliver measurable value. The most successful organizations will be those that:

1. Develop a clear AI strategy aligned with broader business objectives
2. Build strong data foundations to fuel intelligent systems
3. Create balanced technical teams combining specialized expertise with domain knowledge
4. Implement thoughtfully with a focus on change management and adoption
5. Measure comprehensively to capture both immediate and strategic benefits

By approaching AI as a transformative business capability rather than merely a technical project, organizations can create sustainable competitive advantages while preparing for a future where intelligence is embedded throughout their operations and offerings.

Remember that AI implementation is not a destination but a journey—one that requires ongoing adaptation, learning, and refinement as both the technology and business needs evolve.

This article was written by Nguyen Tuan Si, an artificial intelligence strategy specialist with experience helping organizations develop and implement effective AI solutions across various industries.