Updates, announcements, and deep dives from the neww.ai team.
A deep dive into the engine registry architecture that lets us ship a new vertical every week — from search to finance to code agents — all sharing one data layer and auth system.
Traditional search returns links. Ours returns answers. See how we combined Meilisearch with LLM re-ranking to build a search engine that actually understands what you need.
Meet Claw — a sandboxed code-agent runtime with 10 adapters, streaming execution, and first-class Prisma integration. Here's why we built it and how you can use it today.
The average company uses 110+ SaaS tools. We think that number should be one. Here's our thesis on vertical AI consolidation and what it means for the future of software.
Most AI shopping tools just wrap affiliate links. We built ours on real-time crawl data, structured product graphs, and preference learning. Here's the technical breakdown.
One year ago we wrote the first line of code. Today we have 30 products, 53 specialized engines, and a thesis that AI should be one platform — not a thousand point solutions.
We process millions of documents across our products. Here's what we learned about chunking strategies, embedding models, and hybrid retrieval that the benchmarks don't tell you.
53 repositories and counting. We believe AI infrastructure should be transparent and auditable. Here's our engineering philosophy and what we've released so far.