Navaneeth Krishnan

@mscode07 Congrats , lfg!!

I made the first sale 🥳 And it's not even launched yet!!!!

@jesslivingston The worst part is getting attached to the possibility of the outcome. haven't done much fundraising but from whatever rejections i got, i realised this: stay detached from the outcome but don't stop moving

For many startups, fundraising is a grind filled with rejections and occasional humiliations. It can be even more demoralizing when you think you're the only one struggling. You aren't.

@neembu_paani31 Pretty sure they are just restoring the capabilities of Opus 4.6 and repacking it as Opus 4.8 or something

i can smell anthropic is about to release a new model

@kunalvg Controversial take: India needs capitalism so badly. Also, make foreign investment easy. I've literally heard people who badly want to invest in Indian startups but hesitate because of strict regulations

India needs urgent reforms. Reset everything. Now or never. > Push nuclear and solar at unprecedented pace. > Cut freebies and run real welfare instead. > Privatise everything that bleeds taxpayer money. > Finish what was started with agriculture reforms. Time to stop pretending we have time.

@theonkartwt lfg!

Vacation going great❤️

Working in this field, I've figured out a few things: Rule 1: Buy their max plans when they roll out a new version (more internal usage), for the first 3-4 months. Then kill it. They'll probably degrade it. Rule 2: If you hear a competitor is shipping, wait for the benchmarks before continuing your provider plan. Rule 3: Always maintain a memory system so it's easy for you to migrate between providers. Rule 4: Don't be loyal to any provider. (most important) . let them see the churn so they ship fast

Am I the only one seeing this? Three months ago my feed was “delete codex, switch to claude” Now it’s “delete claude, move to codex” Grabbing popcorn. who ships next? mythos? codex 6? what do you think?

@PriyankaSaini28 vibe train any ai model ( private beta) waitlist: railcompute.com

What are you building currently? Share the link. Let's promote it.

@HowDevelop @MiniMax_AI @kubeflow Congrats, lets connect

Recently got a reach out from @MiniMax_AI careers team purely because of my contributions around @kubeflow. No applications. No cold outreach. Contributing to Open Source is probably the highest ROI career move most developers still underestimate. Your GitHub can literally become your resume, portfolio, network, and inbound recruiter pipeline at the same time. The internet notices people who consistently build in public. P.S. The Minimax team is hiring for interns / FTEs for their AI Infra team in Shanghai. In case anyone is interested feel free to reach out! Happy to forward the application!

@Shalini70856041 congratulations

Hello everyone, I am Shalini. Last summer I got the opportunity to intern at Google and also secured a pre-placement offer from there. I usually get a lot of DMs asking how I made it, so I thought of sharing a few things that might help someone. Though last time the PPO conversion rate was quite low, especially in the Cloud org, these are some things that really helped me: Document everything from Day 1. You should always know where your time is going, and you can also share updates with your buddy/co-host/host. Communicate regularly with your host, co-host, or manager. Take feedback seriously and work on it consistently. Write more production-level code and learn company coding standards. While reading existing codebases, observe how engineers structure code, name variables, write tests, and maintain readability. Every company has its own coding style and practices, so learning to adapt to them is very important. Maintain all important documents like design docs, testing docs, and other technical docs. These artifacts are valued a lot at Google and are useful in other companies too. There might be times when you feel overwhelmed or think you won’t make it. Please don’t give up during that phase. Talk to your buddy or teammates — they genuinely help. Don’t ask questions without trying first. Visibility matters, but before asking anything, check available resources and try to figure things out yourself. Don’t compare your team with others. I know team experiences can differ a lot, and comparison can mentally affect you — it happened with me too. Maybe your friend got a “better” team, but comparison won’t help your growth anyway. Lastly, even after giving your 100%, it’s possible to not get a PPO because many things depend on business requirements as well. So after your internship, make sure to keep applying and explore other opportunities too. I hope all of you make it. Best of luck ❤️

@srishticodes depends

As a dev what's your favorite localhost port ?

@nico_jeannen most of their tools stick to one keyword (you liked/engaged/wrote) and assume a lot of things about you which is 100% wrong.

Tried using LinkedIn for 5 minutes yesterday, and got DMs like this Received at least 5 similar ones, I don't understand anything to what those ppl are talking about Am I just retarded or is this how they talk there?

@StevBuilds RailCompute lets anyone train or fine-tune AI models with just natural language, no configs, no YAML, no GPU babysitting. waitlist: railcompute.com

Which problem is your product solving? One sentence + Link 👇

@lets_dig_deeper @rumik_ai Is this open source? happy to contribute in any way possible. Loved this

an audio model that can switch languages, accents, tone, gender, emotions in a single instance this is silk mulberry 1.5 our most cost efficient model! by @rumik_ai research lab

Opus 4.7 has become lazy. Technically speaking, I guess they've limited the verification loop that 4.6 used to do when it launched. That's why I follow a structure to avoid trashy output. In max mode, I always ask it to output a confidence score for the implementation it made, based on code consistency and working capacity. I also make sure to say implementations should be temporary patches rather than permanent fixes. For any implementation where you feel the confidence is less than 90 percent, reject it. If you feel it's still needed, you can proceed to human-in-the-loop (me) and address the issue. This works for me and has reduced the trashy output it gives me

i asked claude opus 4.7 to refactor a large codebase. 68 minutes, millions of tokens burned - it finished nothing worked. app completely broken

@KaiXCreator I am loyal to neither.

Are you team Claude or Codex?

@supabase #do not touch this line

What's the last line of code you wrote without AI?

@adxtyahq I loved the chair man. Keep going

spent the day fixing tests and preparing for a new system architecture discussion + approval

@romanbuildsaas What are you guys building?

At our hacker house in SF, where I live and work with my two cofounders, one of my cofounder’s girlfriends came by recently. After seeing us work 18 hours a day for weeks straight, she said that by the end of YC we’d probably need a full reintroduction program back into normal life… like animals being released from a zoo 😭

@GeekyVaishnavi @Windows @Apple When brothers apply the sisters' skincare pack

.@Windows got that @Apple skincare routine.

@StevBuilds @CesareDadamo sure

@navaneethvb @CesareDadamo Thanks mate! Let's all connect?

Special day for @CesareDadamo and myself! Our first product made more than $25.000💰 And it's just the beginning🔥 Ship today guys!