Matan-Paul Shetrit, Director of Product at Writer, joins Mang-Git to discuss the evolving AI landscape. Matan shares insights from his career in tech, highlighting his time at Flexport, Brex, and A-Team, and delves deep into Writer's offerings including AI Studio, no-code applications, and developer APIs. They explore the build vs. buy dilemma and the real costs and benefits of fine-tuning AI models versus using Retrieval-Augmented Generation (RAG). Matan also touches on the challenges and opportunities of integrating AI into large enterprises, and the importance of tailoring solutions to both technical and non-technical users.
Matan-Paul Shetrit, Director of Product at Writer (www.writer.com), joins Mang-Git to discuss the evolving AI landscape. Matan shares insights from his career in tech, highlighting his time at Flexport, Brex, and A-Team, and delves deep into Writer's offerings including AI Studio, no-code applications, and developer APIs. They explore the build vs. buy dilemma and the real costs and benefits of fine-tuning AI models versus using Retrieval-Augmented Generation (RAG). Matan also touches on the challenges and opportunities of integrating AI into large enterprises, and the importance of tailoring solutions to both technical and non-technical users.
00:00 Introduction and Guest Welcome
01:43 Guest Background and Experience
04:14 Writer AI's Evolution and Offerings
06:48 Enterprise Solutions and Custom Applications
13:03 Build vs. Buy: Personal Insights
15:06 Challenges and Lessons from Flexport
22:51 Brex and Marqeta: Strategic Decisions
26:21 Navigating Spend Management Tools
26:35 AI Models: Build In-House or Use Vendors?
29:15 The Debate on Fine-Tuning AI Models
32:49 Cost Implications of Training AI Models
36:41 Balancing Simplicity and Usefulness in AI
41:52 Personal and Professional Reflections
45:04 Product Recommendations and Final Thoughts