Episode
EP 143 - AI Tools: Using AI Securely with Eva Gao
- Published
- Dec 10, 2025
- Duration seconds
- 3156
- Processing state
not_requested
Actions
POST https://stenobird.com/v1/public/podcasts/product-for-product-management-3487998/episodes/ep-143-ai-tools-using-ai-securely-with-eva-gao/transcription-requests
Idempotently request low-priority transcript generation for this episode.GET https://stenobird.com/podcast/product-for-product-management-3487998/ep-143-ai-tools-using-ai-securely-with-eva-gao.md
Read the agent-friendly Markdown representation of this episode resource.
Summary
We’re thrilled to welcome back Eva Hongyan Gao, Head of Product ESG at AMCS Group, a returning guest (episode 102) and a product leader in B2B SaaS, circular economy, and ESG, for a special episode on using LLMs securely inside the enterprise. Eva joins Matt and Moshe to offer a candid, hands-on look at how AI fits into enterprise toolkits, the challenges of data compliance, and the realities of integrating tools like Microsoft Copilot Studio within strict security frameworks. Eva brings deep experience building for demanding enterprise customers, where success is measured not just by innovation, but by strict ISO, SOC 2, and GDPR compliance. She shares what happens behind the scenes as product leaders and IT teams try to balance innovation, cost, and data protection, sometimes losing sleep over responsible tool usage and ever-climbing AI integration costs. Join Matt, Moshe, and Eva as they explore: Using AI tools in highly regulated, security-conscious B2B enterprise settings The compliance process: from ISO and SOC2 to GDPR and internal AI guidelines Why Microsoft Copilot is becoming the default LLM in enterprises, and what you still need to watch out for Building internal agents and chat interfaces to answer roadmap questions and handle stakeholder requests Lessons learned moving from over-engineered platforms to simpler, compliant AI tools Creative AI workflows, including removing branded assets between Copilot and Figma and orchestrating information for various departments The ongoing struggle: data redaction, internal transparency, and the limits of controlling generative models LLM orchestration: mixing old-school logic with new AI capabilities, and knowing when not to use AI Security best practices and the importance of a trust-based compliance mindset across t…