Blog
Long-form thinking on engineering, agentic systems, RAG, MVPs, and the practice of shipping software.
Multi-modal demos are everywhere. Multi-modal in production has a different set of constraints. A field guide.
RAG demos are everywhere; reliable RAG is rare. The parts that actually decide whether your retrieval works in production.
No-code MVPs are great for validation. They're terrible at scale. The hidden costs nobody mentions when you're picking the platform.
Companion to our Agentic Development 2026 piece. The 3 patterns that earn their keep — and the 3 we've seen kill projects.
The three questions that catch over-engineering before it eats your runway. A field-tested framework from shipping with founders.
Grounding is the difference between an LLM that hallucinates and one you can ship. Here's what it actually means in 2026 and how to wire it into your product.
Voice agents went from gimmick to retention driver. A look at what changed and how to ship one customers actually prefer.
What separates the voice agents people accept from the ones they hang up on. A 2026 stack guide for shipping voice CX that doesn't trigger fight-or-flight.
Who's selling what, who's actually shipping, and what founders should look for when hiring an AI partner in India today.
