Techie. Building in AI. Impatient. High-sensitivity BS sensor. Been there, scaled that. Mentos Zindagi! ⚪️⚪️⚪️⚪️ 🏏 🎾
Bengaluru, India
Joined August 2009
- Tweets7.7K
- Following1.7K
- Followers853
- Likes3.3K
Pinned Tweet
Why urgency?
Because "kal kar lenge" is a very expensive sentence. Do it today, and tomorrow is free suddenly for the next shot.
Double the shots, same 24 hours. That's the whole cheat code!
Prashant Parashar retweeted
Love the pitch.
IMO laziness and attention deficit is not going to get you to your next gig, especially when jobs get scarcer.
Neither will “dear agent, please apply to 100k jobs on my behalf”
Nor will “I need to be top of Hiring manager’s inbox”
It’s going to be a combination of being substantial, being differentiated, telling your story right, and the right matchmaking.
Hi, it's Darshita. I am a designer.
And I'm pitching myself in public 'cause I'm done with a year full of anxiety.
Everyone says design is dead and AI killed it. But the real parasite was hiding in plain sight.
So here's my film, DESIGN IS DEAD. She's bold, fearless and very PINK 💖
Hiring? My DMs are open.
Prashant Parashar retweeted
Sometimes a demo is simply to remind the world that your competition has no desire to dream -- and no ability to evolve.
Prashant Parashar retweeted
Major cheat code: walking around like life is rigged in your favor actually rigs it in your favor
Prashant Parashar retweeted
One question founders always ask when we speak at YC is: How much of the standard startup advice still applies in the age of AI? At this point, nearly all of it. I can't promise that will be true in 10 years, but at this moment, starting a startup is much the same as always.
Prashant Parashar retweeted
It took Alexandr Wang a full year just to get his first meeting with Zuckerberg — and the story of what happened after explains a lot about how Mark actually operates.
"My memory is that I think it took, like, one year to schedule that first meeting."
This was during a period when Mark was fielding an avalanche of AI interest. Scale AI was deep in the space. The intro came through. And then nothing for twelve months.
Once they did connect, Wang started getting something unexpected: long, rigorous WhatsApp responses.
The obvious read is that Zuckerberg is just a busy person with a hard calendar to crack. The more interesting read: the year-long wait and the long WhatsApp replies are the same trait expressed differently.
The CEO who takes a year to schedule a meeting but fires back a rigorous essay on WhatsApp is not disorganized. He's ruthlessly sorted what deserves his best thinking and what doesn't. The delay wasn't neglect. It was prioritization at work.
Prashant Parashar retweeted
If there’s a single word in your AI-generated deck that you can’t explain, don’t share it with anyone.
Prashant Parashar retweeted
Looking back, it is so interesting that we all rushed in 2023-24 to learn LangChain and many other frameworks. And now, almost all major coding harnesses and agents are just hand-rolled loops.
Turns out, the abstraction was the overhead.
Prashant Parashar retweeted
.@OpenRouter co-founder Alex Atallah on the "everyone is building the same thing" take:
"This reminds me of this tweet I saw... Everybody's building an agent loop with notifications, third-party connectors, context management, memory, sandboxes, agentic web search, an always-on agent on top."
"This product is showing up everywhere. It is showing up everywhere, but these are just the new table-stakes primitives."
"A 2005 version of that tweet would be: 'Oh, everybody's building the same thing. A database, a users table, a sign-in page, a sign-up page, a profile page, a logout page. Everything's the same.'"
"There's like a lot of differentiation really. There's table-stakes needs for AI just like there's table-stakes needs for the web."
@alexatallah @amasad
.@OpenRouter co-founder Alex Atallah, in his first podcast since Stripe acquired the company, joins @Replit co-founder Amjad Masad and a16z's Erik Torenberg on why the future of AI is independence and specialization.
In this conversation, Alex walks through how the Stripe deal unfolded, why he wasn't originally looking to sell, and why "payments and inference are going to blend together."
Pre-OpenRouter, the typical AI workflow had one model provider to choose from, and little pressure on that provider to lower prices. Now enterprises are diversifying across labs and open-weight models, and every board is asking about AI costs and benchmarks.
Amjad argues if your company depends on one AI lab, it can turn into your competitor. So Replit is building the layer that lets enterprises use any model and any cloud, without being locked into either.
Alex and Amjad are split on personal agents – Amjad runs one agent across his whole company and loves the cross-domain joins, while Alex says general agents cause you to sacrifice understanding, and argues 10 specialized chiefs of staff beats one superagent.
0:45 How the Stripe deal unfolded
5:05 Why mixing models beats one model
7:25 Forcing the labs to compete on price
8:50 Enterprises want open-weight models
10:30 Every board asks about AI every month
12:25 Why companies must own their intelligence
14:15 Replit as the independence layer
15:10 Everyone is building the same agent
16:35 Why Amjad built bring-your-own-cloud
18:10 Amjad's agent that runs his whole company
19:55 Why 10 specialized agents beat one
23:35 Machines, not humans, should specialize
27:30 Guardrails for agents talking to agents
31:10 Models training their own replacements
33:45 Most tasks don't need a frontier model
40:50 Training small models on Qwen 8B
43:25 The Rust cycle is coming for AI
45:10 Fusion models: frontier quality at half the cost
YouTube: youtu.be/ekK8urKHPMQ
@alexatallah @OpenRouter @amasad @eriktorenberg
Prashant Parashar retweeted
Parag Agrawal was asked to share insights about Elon Musk that the public might not know.
"I have lots of disagreements with him. But I'll share what I think is for founders here, which is what I think the thing you can admire about him.
The urgency and the ability to compress time. And I think having unreasonable expectations of people is mostly a good thing because most people don't understand what they're capable of and kind of implicitly sandbag themselves and implicitly set lower expectations of themselves than they're capable of.
So when simultaneously inspired and pushed with urgency, people can do more than they thought. And I think he can sometimes extract that from people. And when that works, it's powerful."
I too agree with parag elon knows how to compress time and do work in 1 years which would take 5 to 10 years. What your thoughts on these ?
Prashant Parashar retweeted
People seem shocked that the author of Clean Code doesn't read code anymore. But that should not be surprising.
Why did I write Clean Code? What is the goal of keeping code clean? To get it out of the way of the real job -- thinking.
Prashant Parashar retweeted
The best thing you can bring to the table is clarity. Even more so in the age of AI.
Prashant Parashar retweeted
It takes a while to build the confidence, but there's no future where you're manually reviewing every line of agent code. Not much acceleration in that. You need adversarial agent reviews, you need automated testing, and maybe you spot check. That's it. From prompt to production!
Prashant Parashar retweeted
I wrote for @htTweets (Hindustan Times) about why resumes miss the context that often determines whether someone is right for a role.
How a person makes decisions and approaches hard problems rarely fits on a page. A resume records where someone worked and what they achieved, but the judgment behind those outcomes usually stays with the people who saw the work up close.
AI can make that context portable, helping professionals be understood by organisations beyond their immediate networks.
You can read the full story here: shorturl.at/uoLrj
🤖 Made with AI
Prashant Parashar retweeted
my current theory about all the growth roles suddenly opening up at wannabe startups is that there is actually a lot of following the herd mentality; the only way out is either you should hire a top person and pay them top salary or hire a young kid who is in the trenches
Prashant Parashar retweeted
The absolute first tip for anyone looking out.
Don’t give into the temptation of writing “Open to work” on your head on LinkedIn
The L1 signal for the innocents - I am open to a new opportunity, duh
The L2 signal that recruiters receive - What a desperate person.
However good your resume, credentials, actual work and your fit to the job - you are reducing your chances of getting hired massively if you put that green banner on your head
You definitely increase your noticeability. But at the same time you reduce your value
Prashant Parashar retweeted
as we use more ai to write code, and I keep encountering worlds where even if you put all the people on the team together in a room you won't have 100% "understanding coverage" (like test coverage) of the codebase, I keep going back to the Peter Naur piece of programming as theory building.
i'll write more about this - foggy thoughts - but I see the "code is like assembly" future. it is inevitable
when I started getting into moderately sized codebases first time in my life, most teams typically had 100% "code reading coverage" summed up across the team, not necessarily a single person. but for every part of the codebase someone in the team knew exactly what it did
what i have seen in larger codebases (long running oss projects like linux kernel, or internal codebases of bigtech companies), is that because of the team churn, and sprawl of the codebase, you actually can find large swathes of codebase which no one in the current team has it entirely in their head.
it didn't exactly stop these software projects from surviving or thriving or expanding further. what was different though is that regardless of shitty/elegant, the code was written by humans and do a degree hence readable by humans, so even if there are pockets of code no one knows, someone could sit with it for a couple of days and spool it up in their head
ai changes a lot of it. the speed and volume is entirely different. we will have to 'shift up a plane' of abstraction in terms of where do we want to understand the program
and hence Peter Naur was right that the code itself doesn't matter. the theory of the program is what matters and the team still needs to build the theory of the program in their head. the old ways will not work in the new world
