Hameed حميد Afssari افسری

🚨 AI can boost your team—or break your product. 🎙 In my latest episode of The CTO Show With Mehmet Gonullu, I sat down with Eric M., Product Engineering Director at Work & Co, to unpack how to lead and build secure, human-first products in the age of AI. We talked: ✅ The real risks of over-trusting LLMs (yes, slopjacking is real) ✅ Why AI should support, not replace, your dev team ✅ How to balance speed, security, and stakeholder trust ✅ Building psychological safety in hybrid teams ✅ And the underrated power of just… listening Eric’s been a CISO, an architect, and a mentor—and his insights hit home for any founder, CTO, or operator trying to ship securely and lead with empathy. 🎧 Full episode is now live on all podcast platforms + YouTube 📌 Links in the comments 👇

15

1 Comment

Selin Kocalar and I left MIT to build an AI compliance startup, but most people don't realize that it stemmed from having to go through compliance ourselves.   Here's the real origin story behind Delve and our goal to save humanity 1 billion hours of busywork.    —   While Selin and I were building an AI medical scribe at MIT, we hit a massive roadblock: making the scribe HIPAA-compliant.   It took months of work. We had an entire team of 15 MIT engineers working on it. Spent tens of thousands of dollars.   The whole experience was a headache.   Right after, a bunch of founders started coming to us asking how we got HIPAA compliant. So, we turned into free HIPAA consultants. But soon, we realized that there might be a bigger opportunity here.    An opportunity much bigger than the AI medical scribe we were building.   —   In 2023, we applied to YC. But when we landed our interview, they didn't actually care about our medical scribe at all. They wanted to know about how we'd automated our own compliance process.    That's when it clicked: every team goes through the same painful, expensive, time-consuming experience of trying to get compliant. And with both of our backgrounds in AI research, we had so many ideas on how to automate the whole process. We never looked back after that.   —   Selin and I left MIT and moved to SF with one mission: eliminate manual compliance work with AI.   Not optimize it. Not streamline it. Eliminate it. We set out to build a product with deep end-to-end AI automation, not just a glorified compliance spreadsheet.   —   Now, at Delve, we're hiring exceptional engineers who want to build AI agents that handle compliance work quickly and accurately.   We've since expanded beyond HIPAA to SOC 2, ISO, GDPR, PCI, and more.   We're building the infrastructure that lets innovators focus on what actually matters: building impactful products.   And it's working. We've been able to help 100s of customers get compliant without the sleepless nights or stress headaches (including Bland, 11x, and Lovable).   —   The best problems to solve are the ones everyone else accepts as "just how things work."   And we're just getting started. 

199

15 Comments

Partition Pruning & Global Indexes: Many believe that partition pruning only works with local indexes or full table scans (FTS)—but that's not the full story. In my latest article, I break down how partition pruning actually works and why it's not about indexes, but about TABLE ACCESS operations: 🔗 https://lnkd.in/eTcb_kaB #OracleDB #Partitioning #SQLOptimization #DatabasePerformance

73

4 Comments

I felt Zeynep Tufekci’s talk at NeurIPS 2025 was very thought provoking and motivating at the same time. Through historical examples like printing press and cars she demonstrated how the society tends to use outdated mental models to understand new technologies, focusing on the wrong benchmarks (e.g., speed for cars, AGI for AI) and missing the real transformative aspects—like scale and accessibility. The true impact of technology comes from making things easier and more abundant, which disrupts systems built on scarcity and friction. What are the scale and friction issues the (Gen) AI is solving for masses today and how that will transform how we function? It’s to be seen but I really liked her points about how it is about disrupting “Proof of Work”, “Proof of Authenticity”, “Proof of Expertise”, etc. causing various societal risks. Job loss to AI enabled systems is a potential problem but it could have policy solutions (hard to get agreement on but generally feasible) and perhaps may allow a higher standard of life to everyone. So it’s not a “right nightmare” - but the others are because they could lead to disruption and destabilization of various norms. Very fascinating and highly important problems to tackle!!!

45