Skip to main content

Business Challenges

Scalability​

  • Need Low latency
  • Use parallel processing
  • Load Balancing: distribute user requests across multiple servers
  • Elasticity: Scale resources based on demand
  • What is one of the key challenges in scaling AI systems to handle large volumes of data and high user demand?
    • Real-time data processing

Interpretability and transparency​

  • Crucial for understanding how decisions are made
  • Interpretability: the ability to explain or understand how a ML model makes its decisions.
    • Decision Tracking: steps the AI took to make a decision
    • Issue identification: transparency makes it easer to spot and fix problems.
    • Continuous improvement: the model is refined for better accuracy and reliability
    • User trust
  • Transparency: ensures that the inner workings of an AI model are clear and understandable to all stakeholders
  • Why are interpretability and transparency particularly important in AI models used for healthcare or finance applications?
    • They help stakeholders understand and trust the model’s predictions and decisions.

Improving Model Transparency​

  • LIME (Local Interpretable Model-agnostic Explanations) is a technique that explains the predictions of complex models by approximating them with simpler, interpretable models.
    • In a credit scoring example, LIME can show why a low score was assigned, highlighting factors like income or credit history. This transparency builds trust and ensures fairness.
  • SHAP (SHapley Additive exPlanations) is a method that explains the output of machine learning models by assigning each feature an importance value based on cooperative game theory.

Integration with Existing Systems​

  • Which of the following are challenges businesses face when integrating AI with supply chain management systems? Select all that apply.
    • Employee training required to interpret AI-generated insights -
    • Data scattered across multiple sources and formats
    • High initial costs and scalability concerns

Ethical & Regulatory Compliance​

  • What is one primary ethical or regulatory challenge businesses face when deploying AI solutions like Generative AI?
    • Ensuring compliance with data protection laws like GDPR

Compliance in Action​

Understanding ethics in AI​

  • In this example of a job hiring company:
  • Fairness: AI should provide equal opportunities and unbiased decisions, regardless of gender, race, age, or background.
  • Transparency: Users should clearly understand how and why AI makes decisions.
  • Privacy: Protecting personal data—like resumes or interview recordings—is non-negotiable.
  • Accountability: Clear responsibility must be established for errors or unintended consequences.
  • Safety: Systems should be designed to avoid harm to individuals and society.