antrix's profile picture. CTO @ http://endowus.com | Opinions are my own

Deepak Sarda

@antrix

CTO @ http://endowus.com | Opinions are my own

Pinned

As a startup, we want to buy tools and outsource every possible small item that's not core to our business. So I end up evaluating quite a lot of SaaS for our various chores and workflows. We are by no means unique in doing this and hence, a plethora of services have spun up.. /1


I just realised I haven't used Perplexity in weeks now. Between ChatGPT enabling the Search tool by default, and google .com/ai which I now use regularly, didn't feel the need for Perplexity.


Deepak Sarda reposted

*gets up on soap box* With the announcement of this new "code mode" from Anthropic and Cloudflare, I've gotta rant about LLMs, MCP, and tool-calling for a second Let's all remember where this started LLMs were bad at writing JSON So OpenAI asked us to write good JSON schemas…


Who's reviewing these PRs before merge? After a few such "long horizon" tasks assigned to AI, and seeing how difficult it is to properly review the massive resulting PRs, I'm now back to "one task at a time" work assignment to AI.

I gave codex a markdown to keep track of progress and let it chirp away on a massive linter debt, and it worked all night and fixed around ~6000 linter/type issues. (it would stop but I queued a massive amount of continue's to keep it working) Part of my prompt was to google…

steipete's tweet image. I gave codex a markdown to keep track of progress and let it chirp away on a massive linter debt, and it worked all night and fixed around ~6000 linter/type issues. (it would stop but I queued a massive amount of continue's to keep it working)

Part of my prompt was to google…
steipete's tweet image. I gave codex a markdown to keep track of progress and let it chirp away on a massive linter debt, and it worked all night and fixed around ~6000 linter/type issues. (it would stop but I queued a massive amount of continue's to keep it working)

Part of my prompt was to google…


It's fascinating how all the replies to this assume that being able to spin up a new service very quickly is an unequivocally good thing at _every company"_. I genuinely think for smaller companies, being slow here would be "good friction" against unnecessary service sprawl.

Every company should create an internal benchmark measuring how hard it is to create a simple regional CRUD service and publish it to production with bare minimal privacy, security, compliance, and observability features in place.



The irony of this AI slop response is too much!

That’s a brilliant analogy! You’re absolutely right that the 8-second limit becomes a signature limitation that stands out awkwardly, just like overused words reveal an AI’s fingerprints. The comparison to “delve” is particularly apt because: Short clips are becoming the new…



Many startups today seem to optimize for attention first and then try to convert that attn into PMF. Cluely is obvious example but all those slick million-dollar launch videos are the same playbook. The strongest signal for me is who keeps building after the attention moves on.


At @endowus, we've done secondaries for employees every year for the past five years, which is basically most of the company's lifetime. We even worked with @cartainc to make the process smoother. endowus.com/newsroom/endow…

Founders should take secondaries when they can. But don’t stop there. Let your employees participate too. They’re the ones who made the equity worth something alongside you. Some examples of companies making that happen recently: - linear.app/now/giving-our… -…



I still use deep research regularly! Essentially any time I know something will involve reading more than a few search results, I trigger a deep research. Also previously: x.com/antrix/status/…

kind of interesting that after being widely hailed as some of the best agent usecases at the start of 2025, and spawning a bunch of copycats, core leaders of the Deep Research agents have now left and I'd hazard a 75% probability you haven't run a Deep Research in the past month.…

swyx's tweet image. kind of interesting that after being widely hailed as some of the best agent usecases at the start of 2025, and spawning a bunch of copycats, core leaders of the Deep Research agents have now left and I'd hazard a 75% probability you haven't run a Deep Research in the past month.…


Even if AI keeps advancing, limitless, cheap intelligence isn’t inevitable. 50+ years ago, we had the same promise of limitless, cheap energy with nuclear .. and we turned away.


Deepak Sarda reposted

We’re expanding Claude for Financial Services, with an Excel add-in, new connectors to real-time data and market analytics, and pre-built Agent Skills, including cash flow models and initiating coverage reports.


Deepak Sarda reposted

Some real-talk on cross-region replication: Responsible services (hi!) synchronously replicate data to multiple AZs in a region. Hardly anyone replicates synchronously between regions because the latency penalty is too high. So you're using asynchronous cross-region…


How many postmortems after today's AWS outage will end with the same line: "we should really move out of us-east-1 this time"


Deepak Sarda reposted

The exposure of the world to US equities is at record levels. A stock market correction would have more severe and global consequences as compared to what followed the dot-com crash. The tariff wars and lack of fiscal space compounds the problem. The underlying problem is not…

GitaGopinath's tweet image. The exposure of the world to US equities is at record levels. A stock market correction would have more severe and global consequences as compared to what followed the dot-com crash. The tariff wars and lack of fiscal space compounds the problem. The underlying problem is not…

Vertex AI Search renamed to Agentspace then relaunched as Gemini Enterprise which is available in three SKUs: Gemini Business, Gemini Enterprise Standard, and Gemini Enterprise Plus. Distinct from Gemini included in Google Workspace Standard, Plus or Enterprise Plus. Simple!!

antrix's tweet image. Vertex AI Search renamed to Agentspace then relaunched as Gemini Enterprise which is available in three SKUs: Gemini Business, Gemini Enterprise Standard, and Gemini Enterprise Plus. 

Distinct from Gemini included in Google Workspace Standard, Plus or Enterprise Plus.

Simple!!
antrix's tweet image. Vertex AI Search renamed to Agentspace then relaunched as Gemini Enterprise which is available in three SKUs: Gemini Business, Gemini Enterprise Standard, and Gemini Enterprise Plus. 

Distinct from Gemini included in Google Workspace Standard, Plus or Enterprise Plus.

Simple!!

United States Trends

Loading...

Something went wrong.


Something went wrong.