Aishwarya Naresh Reganti

The AI Agent Stack in 2026

The article describes how the AI agent landscape has evolved from a linear "stack" in 2025 to a more complex "operating system" model in 2026. This shift is driven by agents moving…

Problem Comes First: Why the Best AI Demos Don't Start With AI

This article argues that successful AI projects begin with identifying a *problem* first, rather than starting with the AI technology itself. The authors advocate for a "CC/CD loop…

Chai & AI Gems: Where VCs Are Actually Placing Bets in AI (and Where They're Not)

This article summarizes a conversation with Geodesic Capital partner Arvind Ayyala about the current AI investment landscape. Arvind highlights that successful AI startups aren’t n…

How Are People Using OpenClaw?

OpenClaw is a locally-run AI agent that connects to messaging apps like WhatsApp and Telegram, offering a more integrated and persistent AI experience than browser-based assistants…

Evals Are NOT All You Need

This article argues that focusing solely on “evals” (traditional AI model evaluation) is insufficient for ensuring AI product quality. Instead, teams should build a continuous “fly…

Chai & AI Gems: January Edition

This article summarizes discussions from a community of AI practitioners, revealing a shift from exploring *if* to implement AI, to tackling the challenges of *how* to successfully…

The State of Applied AI 2025 Report

The “State of Applied AI 2025” report highlights that 2025 saw a shift from focusing on AI model benchmarks to building *reliable* AI systems that function effectively in real-worl…

What Building AI in Regulated Industries Actually Looks Like

This article summarizes a discussion with André Kerr, Director of AI Product at Amex, about implementing AI within a highly regulated financial industry. Key takeaways include prio…

What Product Leaders at Netflix and Meta Actually Think About AI Product Management

This article summarizes a discussion with product leaders from Netflix and Meta about the emerging role of "AI Product Manager." They differentiate between products *with* AI featu…

From 6,000 Lines of Code in Two Days to Getting Acquired by IBM- The Langflow Story

Langflow, a rapidly growing open-source low-code AI tool, was recently acquired by IBM through DataStax, despite being built by a small, self-funded team in Brazil. The founders le…

How to Choose the Right AI Models For Your Use-Case?

This article argues that choosing the right AI model isn’t a simple decision, but a multi-dimensional process involving enterprise constraints, model type, and use-case specifics. …

Day 10: AI Agent Lessons and What’s Ahead

This article summarizes a 10-day series on building agentic AI applications, highlighting that successful implementation hinges on observability (tracking agent actions) and rigoro…