Balancing Innovation and Responsibility : Anindita Mahapatra
summary
Enterprise blueprint for GenAI that pairs platform-level governance with real-time moderation, using Databricks Lakehouse and Unity Catalog to govern data, models, indices and lineage while an AI Gateway logs requests, PII and model usage. Implements task-specific agents for multi-step workflows (returns, underwriting) and measures quality with automated LLM judges plus human grading. Covers mitigation of hallucination, bias and GDPR/EU AI Act risks, evaluation datasets and traceable deployment criteria so agents can be safely promoted to production.
AI-written summary · about
chapters (16)
- 0:00Balancing innovation and responsibility
- 1:12Generative AI in enterprises
- 2:37Accelerating AI adoption
- 4:01Enterprise AI challenges
- 5:27Ethical risks and compliance
- 7:11AI governance versus moderation
- 8:58Agentic frameworks and applications
- 10:23Databricks data intelligence platform
- 12:30Generative AI architectures
- 15:21Deploying agent systems
- 17:08Governance with Unity Catalog
- 19:55Guardrails and usage tracking
- 20:39Deepfakes and regulatory compliance
- 22:23Global AI governance future
- 22:44Demo of Databricks and Unity Catalog
- 25:39Summary
related talks
11mOpen source GenAI vs. enterprise GenAI: Navigating the future of AI innovation
30mAgentic AI in Action: From Framework to Deployment
39mBuilding a Responsible AI Governance Program
50mOh, the Possibilities: Balancing Innovation and Risk with Generative AI
24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…
40mCode is Life - Building a Responsible Generative AI Solution
26mResponsible AI-What, Why, How and Future!
38mBuilding AI responsibly with a GRC strategy