Azhar

@_dylanga @tryramp Which models are you guys using at inspect? Would love to hear more on this.

In the past week, 60% of all @tryramp PRs were done through Inspect. It’s highly effective and has a very high adoption rate across the entire company (80% for all employees, practically 100% for engineering). It keeps getting better every day as we add more tools for all departments, not just engineering. Sales, product, design, data, customer experience, etc all benefit from Inspect.

@trq212 Our entire Claude organization was suspended without any explanation. We have not violated any terms of service and, in fact, added a few new members to our Premium seats today. Could you please help us resolve this issue? It seems unreasonable that I appealed yesterday regarding a few team members’ accounts, and today the entire organization plan has been cancelled.

@arvidkahl True, agents are going to use the software not humans in a few months/years.

Every SaaS must offer an API to its data & functionality. If you run a SaaS, consider this: eventually, all your serious users will run some kind of agentic tool to get work done. The question is: will you make it easier or harder for them (= their agents) to use your service?

@bcherny This is good one. /loop find all priority issues of customers, create PRDs, validate and ask member to review. Going to use this one in our PM tool.

Released today: /loop /loop is a powerful new way to schedule recurring tasks, for up to 3 days at a time eg. “/loop babysit all my PRs. Auto-fix build issues and when comments come in, use a worktree agent to fix them” eg. “/loop every morning use the Slack MCP to give me a summary of top posts I was tagged in” Let us know what you think!

@noahkagan Sent you a DM.

You build it. I promote it. Building has never been easier. Getting customers has never been harder. So I’m picking 2 products to promote to our 1,000,000+ entrepreneurs for our 16-year anniversary next Tuesday. Drop your product below. Don’t have one? Build it this weekend and come back.

@codyschneiderxx That’s what exactly why we built Contentpen.ai to scale our content operations at ContentStudio.

i can't express to you how stupidly powerful claude code is for SEO when you make .env file containing your: - keywords everywhere API key - your dataforseo API key - data warehouse for google search console data - access to your CMS to publish content some things you can do for your saas company map keyword universe for your brand pull your full keyword universe using keywords everywhere's related keywords and people also search for endpoints then send that entire list to dataforseo's SERP API to see who's actually ranking and where the gaps are. full content calendar with clustering and prioritization generate programmatic landing pages at scale if you're a saas serving 20 industries you hit keywords everywhere for the long tail variations per vertical then check dataforseo for difficulty and SERP features on each one and claude code just generates unique pages with the right semantic terms and schema markup already baked in link building link building using dataforseo's domain intersection endpoint that shows you every site linking to your competitors but not to you. pull backlink profiles for your top 5 competitors run the intersection find the gap scrape contact info and draft personalized outreach emails referencing the specific page they link to. entire pipeline in 8 minutes internal linking build internal linking maps using keywords everywhere's related keyword data to create topical relevance clusters then have claude code generate the actual linking structure across your site. not random links. real semantic relationships that google rewards content refreshing when you have access to your Google search console data, you can refresh the content really easily. I use the graphed .com MCP to give claude code access to my data. I have to do this because if I just hit the API or try to use an MCP, we run into rate limits or we have pagenation errors, so the data ends up being misrepresentative you can then ask Claude, okay, analyze the top pages, find the keywords that they're ranking for. which keywords are we ranking for that aren't on this page? How could we refresh and improve the content to incorporate these keywords in so that we make this blog content rank more effectively for us? Easily a 20% lift to your current content by doing this analytics and tracking and then to track all this work you connect your Google Analytics 4 and your Google Search Console into graphed .com then track all of your SEO reporting in Graphed .com if you try to just connect directly to Google Analytics 4 or to Google Search Console without a data pipeline and a data warehouse, you're going to have duplication errors, run into API rate limits, and pagination. You have to do this to get accurate data at scale

@goetzcj @AgnoAgi Using this in production at @ContentStudioio. It is powering all of our agentic stuff.

Anyone else feel like they discovered jet fuel after using @AgnoAgi in their projects? Playing with their scheduler tomorrow

@pie6k Usermaven.com

What product would you recommend for app analytics? I’ve tried mixpanel, amplitude and trying posthog now, but all are quite disappointing I need something simple, stable and with greasy DX

@mattpocockuk Great Video Matt.

My Ralph Wiggum breakdown went viral. It's a keep-it-simple-stupid approach to AI coding that lets you ship while you sleep. So here's a full explanation, example code, and demo.

@diptanu @tensorlake How are you different from temporal?

If you’re a distributed systems engineer excited about database-adjacent systems, container infrastructure and cluster schedulers, we’re hiring at @tensorlake. Our team has built systems at HashiCorp, AWS, Netlify, and SurrealDB. Build foundational AI infrastructure with people who care about correctness and scale. Our platform lets developers build agents and workflows on a durable, sandboxed, serverless runtime, handling fan-out, retries, and correctness under failure. Below: Claude Agent SDK powered agentic web crawler running on Tensorlake!

@aidenybai centrifugal.dev

what's the best service to run multiplayer/realtime experiences

@iannuttall They had outage last night, seems like they flipped the switch again.

also noticed this and went back to gpt-5.2 high today opus was writing a lot of bad spaghetti code for simple stuff and repeating logic everywhere!

Turn your ideas into reality with just one prompt! ContentStudio now allows you to make on-brand visuals, videos, and captions in a jiffy. So, no more delay and headache. Type your prompt and see the magic happen. Make your life easier and try generating with AI Studio!!!

@leebetts_ Sounds good. 👌

@AzharDataDrift Want to come on the Bootstrapped Heroes pod to tell your story?

We just crossed $20k MRR with Usermaven — fully bootstrapped. The road here was anything but smooth. We burned money in the wrong channels, chased the wrong personas, built the wrong things, and nearly derailed the product twice. If you're building a SaaS with a small team, here are the 7 mistakes that almost killed us — and the lessons that saved us 👇 1. Burning money on ads that never converted We poured thousands into Google Ads early on, thinking paid search would jumpstart growth. Reality check: - CPCs were brutal - Competing with VC-backed giants was impossible - Conversions were almost zero Paid ads didn’t fix our unclear positioning — they only magnified it. Ads work only when your organic engine already works. Our real growth came from SEO and word of mouth. 2. Building every feature people asked for We were drowning in requests: - Heatmaps - Session recordings - Surveys All tempting… all distracting. These features would've slowly turned us into a generic analytics suite. But Usermaven's mission is clear: Attribution + product insights for marketers, agencies, and brands. The hardest part of product isn’t shipping, it’s saying no to seductive distractions. 3. Own your infra from day 1. Usermaven processes millions of data events daily. On AWS/GCP pricing, our margins would've evaporated. So we took the uncomfortable route: switching to bare metal. The result: - Lower costs - Higher performance - Predictable margins If you're in analytics, cloud convenience becomes a silent killer. Own your infra before it owns you. 4. Undercharging hurt us more than it helped We started with low pricing to “make it easier” for customers to join. But low prices brought: - hobby users - uncommitted trials - constant support demands - high churn When we raised prices 3×, everything improved: - higher retention - better-fit customers - more agencies - more serious marketers Pricing is a filter. Cheap pricing brings cheap commitment. 5. Falling for the “big prospect” trap Several big leads told us: “If you build this one feature, we’ll move everything to Usermaven.” - We trusted them. - We built it. - They ghosted. Every. Single. Time. Never build custom features on a promise. If they won’t pay upfront, it isn’t real. 6. Trying to serve everyone destroyed our clarity We tried targeting: - SaaS teams - PMs - Marketers - Bloggers - Agencies - E-commerce brands It confused our messaging and confused our customers even more. Things changed only when we focused on the people who actually felt attribution pain: performance marketers, growth teams, agencies, and data-focused brands. Positioning is subtraction. Clarity beats versatility. 7. Believing great code was enough As engineers, we tried to avoid sales. We thought a great product would magically sell itself. Wrong. B2B is: - trust - relationships - follow-ups - guidance Once we embraced sales, everything changed, faster deals, clearer feedback, lower churn. If we restarted today, we’d bring sales in on day one. These lessons were expensive, painful, and humbling, but they shaped the business we have today. If you’re building something of your own, I hope this helps you skip a few of the scars.

@rudrank Every single time. lol.

Biggest problem with Claude Code

@evolvee33 🙋‍♂️

The longer I stay on X, the more I notice: This place is full of NPCs. – Auto-generated comments. – Same recycled quotes. – No identity. – No soul. If you're one of the rare humans left, say hi and let's connect.

@StockAlertX About 80% of our ad spend went into Google, and the remaining budget was split between Twitter and Facebook. Google was our main focus, unfortunately none of the channels moved the needle for us.

@AzharDataDrift Huge milestone! What was the biggest channel that actually worked vs the biggest money burn? We burned $5k on Google Ads before realizing Twitter was our goldmine.

@ShipAloneCEO Thanks man!

@AzharDataDrift Huge congrats, 20k MRR is a massive milestone for a bootstrapped team. Love the grit, inspiring for solo founders trying to ship.