I had the incredible opportunity to attend the Microsoft AI Tour, and I want to share some of the key takeaways from the talks I attended.

This blog will be a bit lengthy, but I trust it will be very useful if you’re interested in how artificial intelligence is transforming various industries.

My main notes and findings were:

• AI should be driven by business needs rather than just technology.
• The biggest challenge is the shortage of talent in AI.
• Manufacturing companies are already exploring and experimenting with AI.
• Mexico is one of the countries that will invest the most in AI in the coming years.
• At the end of the blog, I’ll include my notes on the keynote by Satya Nadella, CEO of Microsoft, as a closing to this analysis.

# Alfonso Rodriguez

Director of Product Marketing, Microsoft

• Isolated Data = Obsolete: The volume of data is an investment.
• AI in Manufacturing:
  • 64% of manufacturers are researching AI.
  • 35% are implementing AI in production.
  • 58% plan to increase investment in AI.
• Organizational Challenges:
  • Talent shortage in design and engineering.
  • Insufficient model and data maintenance.
• Future Opportunities: Focus on design, development, and operations.

# Diego Bustos

Chief Data/Analytics Officer, Grupo Bimbo

• Digital Transformation: Essential for business survival.
• Challenges: Adoption and scalability of AI.
• AI Approach: Should stem from business needs, not just technology.
• AGILE Methodology: Continuous improvement and product evolution.
• Prioritization: Return on investment and alignment with global strategies.

# Jose Luis Apodaca

Head of Global Data Science, Cemex

• AI and Digital Transformation: Key to extracting value from data.
• Scalability: Necessary infrastructure to support systems.
• Methodology: Two tracks: OPEX (exploration and pilot) and KAPEX (development and implementation).
• Future Vision: Become a data-driven cognitive company.

# Carlos Herquino

Fabric Product Marketing Lead for Industry

• Current Challenges: Data quality; Microsoft Fabric created for end-to-end data cleansing.
• Transformation through AI:
  • Impact on three pillars: individuals, teams, and companies.
  • Time Enrichment: Improving business processes for employees and customers.
• Requirements for Generative AI Models: Powerful models and AI platforms. Unified and clean data.
• Risks of Non-Unified Data: Data copies and structural inefficiency.
• Steps for Data Unification:
  • Unify data in an open government lake house foundation.
  • Harmonize data for AI and establish real-time connections.
  • Implement a Unified AI Platform.
  • Data Flow: Data → Microsoft Fabric → Azure AI Studio.

# Mike Hulme

General Manager, Azure Digital Apps and Innovation

• Reinventing Applications: All applications will be reimagined with generative AI.
• Application Projections: Expected to see 1 billion logical applications by 2028, defined by AI.
• AI Adoption: Over 80% of companies will use generative AI APIs by 2026 (compared to less than 5% in 2023).
• Growth and ROI: Advantage of strategic disruptors: 250% ROI.
• Opportunity Identification: Expert knowledge and a tech stack that supports Open Standards are crucial.
• Azure AI: Enables integration at all steps of applications.
• Microsoft Trustworthy: A program that ensures data privacy and transparency.
• Development Efficiency: 50% reduction in the time to develop new applications.
• Launching Azure Innovate: Helps identify areas for improvement and migration.

# Paul Yu

Developer Advocate

• Intelligent Applications:
  • Interaction through natural language.
  • Data-driven and personalized.
  • Continuous learning and improvement.
• AI Integration: Facilitating the addition of AI in applications using the cloud and microservices.
• CloudNative Features: Speed, agility, and modern design.
• Resilience: The cloud enhances resilience; adopting DevOps and automation is key.
• AKS (Azure Kubernetes Service): Ideal for deploying applications in the cloud.
• Key Takeaways: Deploy quickly and securely, with rollback capability.

# Dona Sarkar

Chief Troublemaker, Microsoft

• Developers: Expected to train models and create bots with AI.
• This is our third rodeo:
  1. Digitizing data.
  2. Uploading it to the cloud.
  3. Training AI with the data.
• The Reality of AI:
  • Nonsense: Critiques about reliability (unfulfilled promises).
  • Real Business: Non-techs solving real problems; new job roles emerge.
• Language Models:
  • SLM (Small Language Model): Specific and viable for concrete objectives, executable in the cloud and locally.
• Model Evaluation: Compare public metrics and manually evaluate responses.
• Choosing the Model:
  • Use case specificity.
  • Available resources.
  • Deployment environment.
  • Desired performance.

# Satya Nadella

Microsoft CEO

• Investment: $1.3 billion USD over three years for AI and cloud infrastructure in Mexico.
• AI for Good: Focus on health, connectivity, and sustainability challenges.
• SMEs and AI: 57% of Mexican SMEs already use AI for efficiency and decision-making.
• Ethical Innovation: Commitment to the ethical use of AI, inclusive growth, and sustainability.
• Skills Development: Emphasis on training for AI adoption across diverse demographics.