درباره این اپیزود
Host Damien Filiatrault talks with Ivan Lee, founder of Datasaur, about the growing importance of private AI and why companies may need to rethink their reliance on external AI providers. As organizations adopt AI across more business processes, concerns about data privacy, regulatory compliance, costs, and long-term ownership are becoming harder to ignore.
Ivan explains how Datasaur helps organizations build and run AI systems within their own infrastructure, using open-weight models and customized workflows. He shares an example of a 133-step insurance claims process that his team reduced from 65 hours to just five, while preserving human oversight at critical checkpoints.
The conversation also explores the technical decisions behind building private AI, from selecting the right models and frameworks to using RAG, fine-tuning, and custom evaluations. Damien and Ivan discuss why not every task requires a powerful language model, how traditional software can work alongside AI, and why testing against real business requirements matters more than general AI benchmarks.
Ivan also shares why he believes AI agents will increasingly replace traditional chatbot interactions in enterprise environments. Rather than requiring employees to manually prompt an AI tool, agents can work within existing systems, handle repetitive tasks, and involve people only when their input or approval is needed.
For developers, technology leaders, and organizations exploring AI adoption, this episode offers a practical look at what it takes to build reliable, private AI systems and why owning AI capabilities, rather than continually renting them, could become an important competitive advantage.
What you’ll learn
- Why regulated organizations are turning to private AI solutions
- How Datasaur reduced a 133-step insurance claims process from 65 hours to five
- When to use traditional software, smaller AI models, or more powerful LLMs
- Why custom evaluations matter more than public AI benchmarks for enterprise applications
- How AI agents can integrate into existing workflows without relying on chat interfaces
- The differences between open-weight, closed-weight, and open-source AI models
- When prompt engineering, RAG, and fine-tuning make sense for custom AI development
- Why access to quality training data is often a bigger challenge than the training process itself
- How owning AI models and evaluation data can reduce costs and create long-term advantages
Memorable sound bites
“Everything is private by default.”
“We took that process from sixty-five hours down to five hours.”
“Chatbots are the most primitive UI we will ever have for LLMs.”
“It's mostly the difficulty and time it takes to gather that data set in the first place.”
“It doesn't make sense to just continue renting our intelligence.”
Links:
Datasaur Website: https://datasaur.ai/
Get 20% off your first month with Scalable Path: https://www.scalablepath.com/commit
Commit & Push Website: https://www.commit-push.com/
Scalable Path Website: https://www.scalablepath.com/
انگلیسی
ایالات متحده آمریکا
رونوشت 🔗
Are you the producer of this podcast?
Add a podcast transcript
Need Audio-to-Text?
Transcribe with Listen411 in Just 60 Seconds
جستجوی اپیزودهای گذشته
اپیزودهای قبلی Commit & Push را جستجو کن.
قسمت های دیگر در این پادکست
Host Damien Filiatrault welcomes Killian Valkhof—long-time web-dev, educator, and solo founder of Polypane, the “browser made for developers.” They trace Polypane’s journey from a scrappy side-project to a profitable, one-person SaaS, and unpack the power—and quirks—of building tools for developers…
Most AI projects fail. Some never ship. Others ship and implode. So what’s going wrong?In this episode of Commit & Push, Damien sits down with Dan Saffer - Associate Director of Outreach at Carnegie Mellon’s Human-Computer Interaction Institute and author of Microinteractions - to dig into the …
In this episode, host Damien Filiatrault sits down with Robin Hanson—associate professor of economics at George Mason University, former AI researcher at Lockheed & NASA, and author of The Age of EM—to explore what happens when the worlds of software engineering and economics collide.
What you’…
Host Damien Filiatrault welcomes Iddo Gino, founder of RapidAPI (launched at 17, later a unicorn, acquired by Nokia) and now CEO of Datawizz. They trace Rapid’s journey from a GitHub list to a global API marketplace, unpack what sky-high valuations actually change, and debate MCP versus simply fixi…
"Don’t hire the brilliant jerk." That’s just one of the hard truths Zach Goldberg, CTO of Gruntwork and author of The Startup CTO’s Handbook, drops in this episode of Commit & Push.
Zach joins Damien to talk about the mistakes most new CTOs make, why developer productivity is wildly misundersto…
سلب مسئولیت: پادکست و آثار هنری تعبیه شده در این صفحه متعلق به Scalable Path است که متعلق به صاحب آن است و به Listen Notes، Inc وابسته یا تایید نشده است.
ویرایش
از کمک شما برای بروز نگهداشتن پایگاهدادههای پادکست سپاسگزاریم