I don't blog often, but when I do, I buy the domain names. 😜 TheNewMoats.com UnitOfValue.com VCrisks.com TheOnDemandEnterprise.com EvolutionOfCloud.com CastlesInThe.Cloud OpenSourceVsCloudCastles.com TheNewNewMoats.com

One of my favorite things about @niteshiftdev using all the new models AND all the new harness features in a single product: - Claude Code dynamic workflows with Fable ($$$) - Codex /goal with GPT 5.5 (very good for optimizations) - DeepSeek V4 in OpenCode

the first engineer I hired at WarpStream had 1 year of professional experience I hired him anyways because he was smart, hungry, eager to learn, and willing to work a lot today he’s a foundational member of the team and I can’t imagine life without him hire the juniors folks

Chrome extension for website -> Figma Design!

.@TechCrunch wrote about how @niteshiftdev is reducing our customers' dependence on OpenAI and Anthropic. Here's how we think about it: In the last software cycle, your most critical upstream vendor was your cloud provider: AWS, GCP, or Azure. In the AI era, it's your token provider: Anthropic, OpenAI, or perhaps Google. Despite the "supply chain risk" from hyperscalers, most companies were *not* actually multi-cloud. At best, you might have a few side workloads that run on a second cloud provider (e.g. data analytics on BigQuery) But even if you aren't *actually* multi-cloud, when contract renewal comes up, it's very important that your AWS sales rep believes that you *could* be multi-cloud. That gets much harder if – in addition to EC2 and S3 – you're also using Lamba, and CloudWatch, and DynamoDB – differentiated services that are hard to migrate off of. The same will be true for your token provider. If you run your whole company on their managed agent platforms, you are stuck. It'll hurt you financially at contract renewal, and strategically, if and when they start competing with you directly techcrunch.com/2026/06/10/da…

Round 2 of building a business in a market everyone thinks will be eaten by hyperscalers I think we’re going to be OK

New @GreylockVC Change Agents is out today featuring @tuhinone from @baseten We discussed all things inference: - Moving from rented to owned intelligence - How the most advanced AI companies are post-training OSS models to give them more control over performance, runtime requirements, and cost - With agents, inference is evolving to a set of tools for continuous learning vs just "run this model" - Capacity constraints are worse than you think, and nobody is willing to take a hit on capability Full episode below

Excited for what @smehmood is building! Greylock continues to support some of the most interesting AI companies and founders.

this is insanely cool coding only represents a tiny fraction of how devs spend their time so you’ll never actually be able to ship faster until you can close more of the loop, looks like this does a lot of that

We're launching @niteshiftdev – the full-stack cloud for coding agents Verification is the new bottleneck. Software teams can now define their dev environment and verification tools once. Then run any frontier agent in the cloud: Claude Code, Codex, or OpenCode

puffy plays pickleball june 30 in SF, co-hosted with good friends luma.com/the-agent-open

We Parse PDFs We spent 7 figures to put this on billboards throughout SF. I thought long and hard about putting something more creative and whimsical. But then you wouldn’t know what we do. AI agents (and humans) are consuming exponentially more documents as they do real work. They need the best quality document parser to not output garbage on downstream tasks. This is what we do today as a company. If you have any PDFs (or other documents), we parse them :) If you’re around SF in June for one of the following events, come stop by our booths: ✅ Snowflake Summit (this week, Booth 1123) ✅ Databricks Data AI Summit (June 15-18, Booth 137) ✅ AI Engineer World Fair(June 29-July 2, Booth L-G47) You can find us by the same sign we put on our billboards! We Parse PDFs @llama_index

"Design vs code" is a false dichotomy. What we're working towards at Figma: freeform design AND prototyping with code AND shipping to production ... all in one connected, collaborative and visual first platform.

Corinne is a force of nature! 💪🚀

An agent that's fluent in Figma and purpose-built for design. On the canvas, where your team works and design context lives... rolling out in beta starting today!

The only way to be a coding agent company is to be a model company. No other business model is sustainable at scale.

Persistent agent VMs are a fundamentally hard-to-run workload (they’re “pets”, not “cattle”). AFAICT, no provider can run it reliably. Applications are going to have to architect for VM failures, like Claude Managed Agents does.

Palo Alto Networks at @GreylockVC