Expert Panelists (L to R, as in photo):

• Himanshu Jain (IIT Bombay, CVS Health/Aetna)
• Richa Naik (IIT Madras, MathWorks)
• Divi Lohia (IIT Bombay, C3AI)
• Shrikant Kshirsagar (IIT Bombay, Prioriwise). 

The panel was expertly moderated by Durriya Doctor (IIT Bombay, MathWorks), president of IIT AGNE.

Strategic Approaches to GenAI Use Cases

The panel initiated a thought-provoking conversation on how organizations can strategically evaluate Gen AI opportunities and identify use cases to harness its power.

Enhancing Developer Productivity: Richa Naik shared how MathWorks is seamlessly integrating generative AI into MATLAB. This integration leverages MATLAB's deep understanding of developer projects and workflows to provide context-aware intelligent code completion, chatbot-assisted brainstorming, and powerful tools for learning to code and debugging errors more efficiently. Utilizing the bigger context, MATLAB provides a much more powerful solution than third-party LLM plugins.

AI-Driven Project Alignment for SMBs: Shrikant Kshirsagar highlighted how the Prioriwise platform helps small and medium-sized businesses. By offering AI-driven recommendations on high-impact projects, Prioriwise helps these businesses align their strategic goals with their tactical efforts. This is particularly crucial for SMBs with limited resources, ensuring efficient allocation of time, money, and other assets towards their organizational mission.

Four Archetypes of Gen AI Use Cases: Divi Lohia presented an insightful approach to categorize GenAI use cases into four archetypes:

1. Knowledge Assistance: Using Gen AI to aggregate large amounts of data (e.g., manuals, SOPs, maintenance logs) to help inexperienced personnel quickly troubleshoot complex issues in industries like manufacturing.
2. Extraction, Summarization, and Generation: Automating tasks like Request for Information (RFI) generation or contract analysis and checklist creation.
3. Data Analysis: Using Gen AI to analyze both structured and unstructured data, such as benchmarking suppliers by analyzing thousands of incoming invoices.
4. Reasoning: Facilitating complex decision-making, such as generating production plans, identifying shortages, and picking and executing mitigation strategies for supply chain managers.

Enhancing vs. Creating Capabilities: Himanshu Jain offered another perspective for categorizing AI use cases based on the impact they have on the current workflows:

1. Automating current Manual Processes: Replacing human-intensive tasks like reviewing documents or call transcripts.
2. Scaling Existing Processes: Super-charging tasks that were previously limited by human bandwidth, such as analyzing all customer reviews instead of just a small sample.
3. Enabling New Capabilities: Creating entirely new workflows and business value streams that were previously impossible, particularly with unstructured data.

Surprises and Emerging Challenges

Durriya posed a compelling question to the panelists: What has surprised them the most while working with this rapidly evolving technology? Their responses offered invaluable insights.

• The "Plumbing" Problem: All panelists have discovered that immense engineering effort is required to transition a Gen AI application from a proof-of-concept to a scalable, reliable production system. This "plumbing" involves prompt engineering, data integration, creating feedback loops, ensuring security, and establishing robust guardrails against issues like hallucination – often demanding more effort than model fine-tuning itself.
• Unprecedented Pace of Change: Panelists unanimously noted the incredible speed of technological advancement, far exceeding their initial expectations.
• The rise of "Vibe Coding": How low-code/no-code tools powered by GenAI are allowing non-technical users, including high school students, to build working prototypes quickly. The challenge, however, is that users may overestimate these tools and lack the expertise to debug the subtle errors they can introduce.

A Glimpse into the Future

The event ended with a very engaging Q&A session, with questions ranging from the definition of "plumbing" to the societal impact of AI on jobs. One of the most forward-looking questions came from a young student in the audience, who asked how schools can embrace AI for learning while addressing concerns about cheating. The panel agreed that this topic deserves its own future session.

The event concluded with a powerful message from the panelists:

"Gen AI technology is here to stay. Everyone should use it both at work and in their personal life. Treat it as a partner and embrace it or you might be left behind."