Varun Prashar

Co-Founder at Brainspack

[Varun Prashar @Varunprashar]

The future will not belong to people who blindly use AI to build faster.
It will belong to people who can use AI aggressively, but still understand the system deeply enough to take responsibility when it breaks.
That, to me, is the new engneering advantage.
#ai #claude #developer

[Varun Prashar @Varunprashar]

Building the Multi-Agentic AI platform for creating dynamic workflows for creating dynamic agents with different expertise.

[Ahmad Awais @MrAhmadAwais]

I'm close to hitting 50K on X. Should celebrate.

Giving away 3 Pro subs to @CommandCodeAI

STEPS
1. Follow @MrAhmadAwais
2. Follow @CommandCodeAI
3. Repost this post

And reply with what you're your building with Command Code.

Will pick randomly soon as I hit 50K. LFG!

[Varun Prashar @Varunprashar]

@VaibhavSisinty u are too late to discover it bro

[Varun Prashar @Varunprashar]

#openai discovered a completely new geometric structure humans missed for nearly 80 years and it could influence how we think about networks, chip layouts, and spatial optimization.

This is Amazing!

#ai #engineering #llm #developer

[OpenAI @OpenAI]

Today, we share a breakthrough on the planar unit distance problem, a famous open question first posed by Paul Erdős in 1946.

For nearly 80 years, mathematicians believed the best possible solutions looked roughly like square grids.

An OpenAI model has now disproved that belief, discovering an entirely new family of constructions that performs better.

This marks the first time AI has autonomously solved a prominent open problem central to a field of mathematics.

[Varun Prashar @Varunprashar]

@Its_Nova1012 Only plan is to keep building amazing products Which solves real pain & add value to customer.
No matter if AI is doing it efficiently

[Varun Prashar @Varunprashar]

without sub-agents, it would have taken 35-60 mins!

[Varun Prashar @Varunprashar]

#codex Worked for 15m

Subagents used:
Backend/security: test callback, token migration, docs. Market/scheduler/audit: readiness, scheduler history, redaction.
Frontend: module split, modal/chart/status/risk UX.

#ai #developer #genai #chatgpt #openai

[Varun Prashar @Varunprashar]

@elonmusk bro i dont have grok-heavy ..its too expesnive for 8 month old startup!

[Varun Prashar @Varunprashar]

@sama Attention to detail nd better memory understandjng. It should know how i think, what i want and how i want

[Varun Prashar @Varunprashar]

Everyone is chasing better models.
SubQ is chasing context scaling.
12M context + sparse attention + claimed lower inference cost is a serious leap.

If it holds up, long-context AI may enter a new phase.

The win is not bigger context.
It is usable, fast, affordable context.

[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.

[Varun Prashar @Varunprashar]

@VaibhavSisinty Heygen too

[Varun Prashar @Varunprashar]

@gdb agentic AI platform to create confugrable agents

[Kimi.ai @Kimi_Moonshot]

Meet Kimi K2.6 Agent Swarm 👋

Highlights：
🔹 Swarms, elevated - 300 parallel sub-agents × 4,000 steps per run (up from 100 / 1,500 in K2.5).

🔹 Outputs are real files, not chat - one run delivers 100+ files, 100,000-word literature reviews, or 20,000-row datasets.

🔹Heterogeneous skills - search, analysis, coding, long-form writing, and visual generation all running in parallel

🔗Try it at: https://t.co/2Tu8McUaUa

[Varun Prashar @Varunprashar]

@sama Its just that its unpredictable for startups that openai will finish their product anyday as a single release. Like u launched clinicians, configurable agents , rag apps etc.

[Varun Prashar @Varunprashar]

Watch the live hands-on: https://t.co/JeJPyz44Ef

[Varun Prashar @Varunprashar]

Watch the live hands-on: https://t.co/JeJPyz44Ef

[Varun Prashar @Varunprashar]

@AnkurPatel59 Setup A is clear winner here.

[Google Research @GoogleResearch]

Introducing TurboQuant: Our new compression algorithm that reduces LLM key-value cache memory by at least 6x and delivers up to 8x speedup, all with zero accuracy loss, redefining AI efficiency. Read the blog to learn how it achieves these results: https://t.co/CDSQ8HpZoc

[Naval @naval]

The biggest question every company faces right now - “does AI acceleration help me or hurt me?”

[Varun Prashar @Varunprashar]

We’re hiring a 𝐁𝐮𝐬𝐢𝐧𝐞𝐬𝐬 𝐃𝐞𝐯𝐞𝐥𝐨𝐩𝐦𝐞𝐧𝐭 𝐄𝐱𝐞𝐜𝐮𝐭𝐢𝐯𝐞 (𝐁𝐃𝐄)

📍𝐌𝐨𝐡𝐚𝐥𝐢 | ⏳ 0–1 Year (Freshers welcome!)

𝐖𝐡𝐨 𝐬𝐡𝐨𝐮𝐥𝐝 𝐚𝐩𝐩𝐥𝐲:
• Freshers & Early professionals
• Sales-driven & eager to grow

#hiring #sales #LeadGeneration #freshers #BDE