Puspesh retweetledi
Puspesh
670 posts
Puspesh
@puspesh
Hard-to-please product engineer, Day dreamer, 2x Tech Entrepreneur (1x Exit). Into books📕 + podcasts 📻 and learning something new🧑💻
Singapore Katılım Temmuz 2008
219 Takip Edilen151 Takipçiler
@nejatian Very cool. After applying, when you want to see "how" they used AI, we recommend using @fairground_work to conduct AI-native technical rounds
English
Opendoor is hiring AI Ops Engineers in Toronto.
If you apply to this role ONLY using AI (including filling out the forms + creating the documents) & tell us how you did it, we’ll move you straight to final round interviews. Extra points for creativity.
ats.rippling.com/en-CA/opendoor…
English
@charliekerr @privy_io @MetaMask @zerion @LineaBuild @tempo We believe you will be doing AI-assisted interview rounds (sync, async or both) - we at @fairground_work built an interview platform to help with this. fairground.work/hire-ai-ready-…
English
Staff Eng is focused on the trading API. If you've used @privy_io, @MetaMask, or @zerion recently you've used the API.
Sr Eng focuses on routing. We recently launched @LineaBuild + @tempo - 41+ total chains on v3. Routing problems are multi-chain and multi-version (v2, v3, v4)
English
We're hiring @Uniswap. NYC, or willing to move (we'll help). Senior + Staff Backend. Product Design. Brand Design. Platform Eng. Protocol Eng. Data Eng. BD/Partnerships. DM's open.
Role details -> careers.uniswap.org
English
@elmd_ @boltdotnew Would love to help - If you are looking to give sync or async interviews to test AI skills, let us know at @fairground_work .. we have built something to solve for this fairground.work/take-home-earl…
English
We're hiring a Senior/Staff Applied AI Engineer @boltdotnew ⚡️
Come build AI agents that turn natural language into production-ready apps.
Remote, global, no degree required.
Feel free to slide into my DMs!
Links in 🧵
English
@CaimingXiong If you are looking to give sync or async interviews to test AI skills, let us know at @fairground_work .. we have built something to solve for this fairground.work/take-home-earl…
English
We’re hiring exceptional AI engineers, especially people with deep experience in large-scale infrastructure, training systems, and production-grade AI platforms. If you’re interested in joining us, please send your CV to talent@recursive.com
English
Today, we’re excited to launch Recursive (@recursive_si): an exceptional team across London and San Francisco, building AI systems that can safely improve their own capabilities over time.
Recursive@Recursive_SI
English
@thebc12 @harmonic_ai Great. If you are looking to give sync or async interviews to test AI skills, let us know at @fairground_work .. we have built something to solve for this fairground.work/take-home-earl…
English
NYC Engineers - @harmonic_ai is growing and hiring across the following roles:
🏗️ Senior Full Stack Engineer (70/30 backend)
📱 Staff iOS Engineer (founding)
🖼️ Staff Front End Engineer (Infra)
🕵️Staff SWE (Recommendations)
We are located in Flatiron, have PMF and are building some really cool ish.
DM's are open!
English
@ajay_2512x @mem0ai @taranjeetio If you are looking to give sync or async interviews to test AI skills, let us know at @fairground_work .. we have built something to solve for this fairground.work/take-home-earl…
English
Mem0 is hiring 2 Engineers in India with salaries up to ₹1 Crore cash.
Open roles:
• Backend Engineer
• Frontend Engineer
Experience required: 7–10 years.
Remote-first.
Work from home, Goa, the Himalayas , they don’t care where you work from.
No IIT filter.
No LeetCode interviews.
No generic resume screening.
Their hiring focus is simple:
“How do you build with AI?”
Instead of resumes, they want to see real AI-assisted work:
• Claude Code transcripts
• Codex sessions
• Cursor/Windsurf prompt logs
• Or workflows showing how candidates planned, designed, debugged, and shipped a product using AI.
PDF or Markdown submissions are accepted.
Applications without AI work traces will not be reviewed.
They’re prioritizing clear and thoughtful applications and plan to close offers by May 19.
Apply at: india@mem0.ai
English
@sama There should be slow and fast lane with differential pricing. Usecase - for sync flows, I would want to use fast mode but for async work (when my agents are working) I don’t mind slow mode. Same intelligence, lower costs will make users work/plan their work differently
English
i get some anxiety not using the smartest-available model/settings.
but sometimes i dont mind if it's really slow.
i wonder if we should focus more on a price/speed tradeoff relative to a price/intelligence tradeoff.
English
Maybe best way to work with agents would be to not directly work with them - in a sync mode.
Rather delegatemaxx and leverage agents for async work. Collect this. Research that. Draft these. Analyze those.
And then work on those artifacts in sync yourself, without agents.
English
My Mac had less available memory than I expected, turned out the "claude" Claude Code processes on this machine (running in various terminal windows) were consuming ~30GB on their own!
The largest one was using 4.9GB
English
@theguywithideas Slippery slope. Okay till you have a counter metric for feeling good about the week as well ..
English
if my claude max plan usage doesn't cross 90% in a week, the week feels less productive. Am I the only one?
English
Puspesh retweetledi
bro you have to try 4.7. 5.5 is better. totally crushes 4.3. they 100% nerfed 3.8 and it shows. i'm running all my workflows on 6.7 now. you should try 1.0.2.5. just re-run the same prompt on 3.3. just try 7.1. pay for 7.1. hit your usage limit on 7.1
English
@artman @linear If you are hiring engineers, would love for you to give this a try - fairground.work/hire-ai-ready-…
English
Today is a hard day. I shared this note with the @linear team today: We’ve made the difficult decision to increase our workforce. This is not a cost-cutting exercise or a reflection of anyone’s performance. We’re simply reimagining every role for the agentic AI era. We’re hiring. We’re sorry about that.
English
Puspesh retweetledi
Fairground CLI Interviews - available now!
Supports both - human (sync) and take-home (async) modes.
Level up your technical interviews today! Emails us at team@fairground.work to know more.
English
Sometimes you feel vindicated, sometimes impressed and sometimes just angry on repeated mistakes.
I think that the model personalities will have measurable effect on AI-powered efficiency. We are going to see tons of new human problems - AI psychosis is just one of them.
English
What's your favorite LLM/AI for non-coding artifact building? (PDFs, HTMLs, Slide decks, one-pagers, docs).
What do they get right?
What are they yet to get right?
Claude is the king for me.
- It "gets" design, layout, and structure. Almost always on the first attempt.
- Falls short on copy and requires multiple iterations to get it right, and sometimes even then it doesn't.
- Good enough for diagrams. Generates good diagrams that you can take and edit (ask it to create an SVG instead of PNG, 10x better outcome).
English
