Twitxy Saved

Markandey Sharma

Verified account This account is verified. Learn more

@TechByMarkandey

Sharing insights on AI, No-Code, Tech Tools & prompts | AI Enthusiast | Unstop's Top 75 Unstoppable Education & Career Influencer | Top AI Voice
611 Following 65K Followers 53K Posts

Latest posts 5

Pinned post

Markandey Sharma Verified account @TechByMarkandey
Everyone using agents should be scared.

It’s insane how easily this company hacked into frontier models and agents.

Your data isn’t as protected as you think it is.

Check out these major benchmarks before you rely too much on AI.

PromptArmor hacked it so you don’t have to find out the wrong way.
Shankar Krishnan Verified account @PromptArmor 2d
Today we’re releasing FrontierBreach, a new standard for attack success rate on agentic workflows in the enterprise.

Existing benchmarks systematically underrepresent the security risks facing modern AI agents.

FrontierBreach reflects attack scenarios on realistic enterprise workflows, while using sophisticated attack chains that result in cyber consequences (ransomware, data exfiltration, phishing, etc).
View quoted post on X
13 replies 14 reposts 204 likes 43K views View on X
Markandey Sharma Verified account @TechByMarkandey
GTA 6 has everyone talking about game graphics, so I had to zoom in on these.

MiniMax H3 Image packs a ton of detail into 4K — from the cars and buildings to the distant cityscape. Even the smaller elements hold up surprisingly well.

And getting visuals like these in around 10 seconds? With 4K generation at 50% off, that's pretty hard to ignore.

@MiniMax_AI #H3Image
9 replies 2 reposts 9 likes 7.2K views View on X
Markandey Sharma Verified account @TechByMarkandey
Not every task needs an expensive AI model.

The real challenge is knowing which model is right for each task.

That's what Drex Router does. It automatically picks the right model based on the task while balancing answer quality and cost.

And the results are impressive:
- 8.80/10 answer quality
- 28x lower cost
- 87.6% MMLU accuracy

You get almost the same answer quality while spending much less.

I like the idea of making AI agents smarter about choosing models instead of sending every task to the most expensive one.

A practical way to make AI agents more efficient.
Nace AI Verified account @NaceAI 19h
Introducing Drex Router.

A small model that picks the right LLM/SLM and reasoning effort for every step your agent takes.

Accuracy: Drex Router 87.6% vs. Jev Router’s 86.4%

Try it: console.nace.ai

COMMENT DREX and get extra $5 in API credits
View quoted post on X
15 replies 57 reposts 267 likes 35K views View on X
Markandey Sharma Verified account @TechByMarkandey
AI can write code fast, but finding and fixing bugs before production is still a challenge 😳
Roee Adler Verified account @roee 22h
“It looked good to me.” The five most dangerous words in the age of AI-generated code.

Just because it looks right doesn’t mean it’s ready to ship.

Who would’ve thought we’d make a musical about it at @hud_hq?
View quoted post on X
4 replies 19 reposts 72 likes 8.6K views View on X
Markandey Sharma Verified account @TechByMarkandey
Building an AI agent is just the beginning.

Making it work in the real world is the hard part.

You need databases, authentication, sandboxes, background jobs, and more.

Volcano brings all of this together in one platform, so developers can build, run, and deploy agents without connecting multiple services.

A much simpler way to build and scale AI agents.
Marco Palladino Verified account @subnetmarco 1d
At Kong we run half a quadrillion AI tokens a year, and I kept seeing agents stuck on clouds built for websites.

So we built a cloud for agents.

Volcano is live today: the self-healing agent cloud. Sandboxes, durable workflows, Postgres, hosting.
View quoted post on X