Sandy

3.6K posts

Sandy banner
Sandy

Sandy

@ssandeshwar

Building AI-driven experiences in JS, Python, and Rust. Senior Engineer dropping language tips, deep immersion reflections, and networking vibes.

Katılım Ocak 2011
838 Takip Edilen1.3K Takipçiler
Sandy retweetledi
George Munguia
George Munguia@jobsbygeorge·
We are HIRING: Full-Stack Software Engineers at Harmony AI.🦅🦅🦅 We’re scaling our engineering team fast. Our engineers live in factories across the United States. They work on-site, in-person, and deploy AI on the factory floor where the work actually happens. If you’re interested, or know someone great, reach out. We pay a $10,000 referral fee for successful hires. Come build the next AI operating system for American manufacturing. 🇺🇸 JOB DESCRIPTION / APPLY BELOW
George Munguia tweet media
English
28
10
204
20.4K
Sandy
Sandy@ssandeshwar·
@Y0BESH Gave the assessment 3-4 weeks ago, heard nothing yet
English
0
0
0
119
Y0BESH
Y0BESH@Y0BESH·
Data annotation is at its peak season. Apply as a generalist if $25/hr is not less for you. Requirement: 👇
Y0BESH tweet media
English
37
24
297
29.3K
Sandy
Sandy@ssandeshwar·
@BrendanFoody And my application is still unanswered for weeks now
English
0
0
0
609
Brendan (can/do)
Brendan (can/do)@BrendanFoody·
just hired 10,000 people in 4 days
English
62
8
422
53.6K
Sandy retweetledi
Ayush Agarwal
Ayush Agarwal@ayushagarwal·
we're hiring backend engineers at @dodopayments dodo is the billing and payments infrastructure for modern software. we process payments across 190+ countries. local payment methods, tax compliance, subscriptions, webhooks, payouts. all in rust. based in bengaluru. 2-5 years experience. we care less about your resume and more about what you've built. personal projects, open source contributions, side projects you shipped. show us what you've made. we're looking for builders who take ownership and take pride in their work. if that's you, apply. application link in the first reply. share this with someone who'd be a good fit.
Ayush Agarwal tweet media
English
30
20
358
29.7K
Sandy retweetledi
Alexander Whedon
Alexander Whedon@alex_whedon·
Introducing SubQ - a major breakthrough in LLM intelligence. It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA), And the first frontier model with a 12 million token context window which is: - 52x faster than FlashAttention at 1MM tokens - Less than 5% the cost of Opus Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention). Only a small fraction actually matter. @subquadratic finds and focuses only on the ones that do. That's nearly 1,000x less compute and a new way for LLMs to scale.
English
1.5K
2.9K
23.1K
12.6M
Ted Nyman
Ted Nyman@tnm·
now hiring for future of devtools at @openai. dms are open.
English
81
27
1.4K
127.3K
Sandy retweetledi
Sandy
Sandy@ssandeshwar·
I got paid this week just for giving feedback to AI models 👀 @outlier_ai_ lets you work from home, pick your own hours, and the tasks are actually fun. Join using my referral link 👇🏽 app.outlier.ai/expert/referra… My first week earnings $$$
Sandy tweet media
English
2
2
4
353
Sandy retweetledi
WeSee
WeSee@weseeventures·
WeSee is Hiring 🚀 We’re building something exciting and we want passionate people to be part of it Open roles: – Blockchain Developer – Full Stack Developer – Marketing Executive – Content Creator (social media + reels) If you’re driven, creative, and ready to build something impactful we’d love to hear from you DMs open | Reply below | Tag someone who should apply 👇
WeSee tweet media
English
156
21
323
19.9K
송준 Jun Song
송준 Jun Song@jun_song·
Are there any AI harnesses better than Codex right now? Curious to know what makes them better!
English
27
0
39
8.6K
Sandy retweetledi
LinaHua
LinaHua@Linahuaa·
Indian people were more advanced than White people for over 90% of recorded history, according to the maximally truth-seeking Grok. But you never hear Indians talking shit like: "LOL- your greatest conqueror in history (who was genetically not even really related to you) shat his pants after barely defeating one of our random minor kings, and his army morale dropped so hard, it made him cry like a baby" Clinging to temporary historical Ws, and belittling others, leads to hubris. Hubris leads to underestimating the Vietnamese, the Afghans, the Iranians, the Chinese.... I remember when American billionaire Todd Boehly bought Chelsea Football Club and bragged about having far more advanced sports data science than the "backwards" Europeans. Now look at where Chelsea is- it's a super banter club. Stay humble. Learn from everyone.
LinaHua tweet media
English
362
643
3.7K
320K
Hamid
Hamid@Hamidk_12·
@OpenAIDevs NS41... New Series 4.1? Or a secret scaling law? The suspense is real.
English
1
0
7
13.2K
Sandy retweetledi
WeSee
WeSee@weseeventures·
We’re hiring an AI + Full Stack Developer Full-time | Onsite | ₹37 LPA Join us in building intelligent systems that solve real problems and drive impact Apply: hr@weseegpt.com
WeSee tweet media
English
41
8
197
44.2K
Sandy
Sandy@ssandeshwar·
@LottoLabs Building a Rust based browser with it (of course a bare minimum one) and it's doing great. Ran straight for hours, fixed over 90 bugs in the code that it wrote
Sandy tweet media
English
0
0
5
368
Lotto
Lotto@LottoLabs·
Assume qwen 3.6 27b isn’t actually opus level or even sonnet level knowledge It’s much better than sonnet 3.5 level And that was sota not long ago, a loved model We’re in crazy times
English
28
18
585
23.6K
Sandy retweetledi
Ishaan Sehgal
Ishaan Sehgal@ishaansehgal·
we killed our $20/month subscription. @omnaraai is now free. for everyone. and the desktop app just dropped. one command center for ai coding agents across desktop, web, mobile, and watch.
English
66
24
388
59.8K
Sandy retweetledi
Xiaomi MiMo
Xiaomi MiMo@XiaomiMiMo·
Xiaomi MiMo-V2.5 Series: Pushing Open-Source Agents Forward 🔸 MiMo-V2.5-Pro, our strongest model yet. A major leap from MiMo-V2-Pro in general agentic capabilities, complex software engineering, and long-horizon tasks, now matching frontier models like Claude Opus 4.6 and GPT-5.4 across most benchmarks (SWE-bench Pro 57.2, Claw-Eval 63.8, τ3-Bench 72.9). It can autonomously complete professional tasks involving 1,000+ tool calls, work that would take human experts days. Tech Blog: mimo.xiaomi.com/blog/mimo-v2.5… 🔸 MiMo-V2.5, native omnimodal with strong agentic capabilities. Pro-level agent performance at roughly half the cost. Improved multimodal perception across image and video understanding, native 1M-token context window, and significantly more efficient inference. Tech Blog: mimo.xiaomi.com/blog/mimo-v2.5 🔗 API & Token Plan: platform.xiaomimimo.com/token-plan
Xiaomi MiMo tweet media
English
136
277
2.6K
363.3K