
🇮🇳 | Forensicator | Hacker | Mentor | Learner #CloudNative #CloudSecurity #DevSecOps #DigitalForensics #IncidentResponse #DFIR #ThreatIntelligence
India
Joined March 2010
- Tweets3.4K
- Following670
- Followers857
- Likes3.6K
Deep Shankar Yadav retweeted
Replying to @Winter_arc26
Theek hai, done! Once I receive my X payout, I'll get you 3 months of X Premium.
Anybody running voice agents for personal use? tired of people calling me all the time. I'm thinking of forwarding all my calls to the v-agent, where it can ask what they want, then I can decide what to do?
I know voicemail exists but people don't know how to use it.
1
37
@shantanugoel are you or know anyone who is running anything like this?
1
14
How does one hire for marketing roles now ?
A friend in Gurgaon is struggling to find a marketing lead for his cybersecurity shop. It’s an in-office role in Gurgaon, but everybody he has interviewed wants to move to Bangalore. Nobody wants to work in Gurgaon?
40
Calling all the internet backbone ppl out there:
I have an ILL from @TataTeleBiz where I am facing a weird routing issue where suddenly some of our public IPs just disappears from public rt. TataTele is out of options; can someone help me debug this.
2
1
85
Whenever I see someone near me getting confused by too much propaganda, whether from the LW or RW, I recommend they read @RidT’s Active Measures.
The book has been incredibly helpful for people to understand how political disinformation works and encourages them to think objectively rather than emotionally about the content they consume.
126
Finance professionals and lawyers spread the most cyber-related misinformation, and this tweet is a prime example. Juice Jacking, while a practical capability, is difficult to execute and modern phones require user interaction before it can be done.
This is simply fear-mongering.
300
Cyber security industry is one of the most impacted due to AI slop marketing ; There is more security writing being published than ever and less of it is worth the time. I keep getting halfway through a article before realising I've read it already someone took another team's work, reworded it, republished it, and the technical detail didn't survive the rewrite.
So a few months back, I built agents reading 130+ sources into a private Telegram channel just for me. The whole setup was built with one goal and that was give me actual security signals: less marketing, less noise, and fast enough that I'd see something before it turned into somebody's collateral. @karpathy LLM Council gave me the shape for it
The agents fetch full articles every few hours, because you can't judge substance from a two-sentence summary. A model drops the obvious chaff, then models from three different labs read every surviving candidate independently and vote. Given the same rubric and the same article, they grade in consistently different ways. They agree on the verdict about nine times in ten; where they split, an arbiter reads the dissent and decides.
However, They aren't ranking importance, they're answering one question only does the article actually presents something or is it Marketing dressed as research, coverage of coverage and generated filler get dropped. Original research, primary sources, real advisories and techniques you can apply get kept. The same story from six outlets folds into one entry with credit to the rest, and where an article publishes indicators, they're extracted and listed alongside it.
I always felt a little selfish keeping it private, because plenty of people have this same problem. Watching @arpit_bhayani's TDD and @shantanugoel's grep gave me the push to take it out of a Telegram bot and put it on the open web for everyone and It became a weekend project, and an excuse to finally use two domains I've been hoarding for years.
today.cyberfortnightly.com - the rolling feed, updated through the day
cyberfortnightly.com - the fortnightly dispatch, best of the last 15 days; first issue on 21 August
threatresearch.io - core research only
Free to read, no signup. If it gets something wrong, tell me and I'll fix it. If you have some improvement ideas, do share those as well.
If you have some good cyber/info sec sources that I am not watching, sent it over.
Is there any decent alternative for @bolna_dev, stuck on a problem but it seems the support is not existing? I have emailed them and also raised queries via ChatBot but never got any response.
35
We recently identified a security issue involving the third-party developer library Axios that was part of a broader industry incident. We found no evidence that OpenAI user data was accessed, that our systems were compromised, or that our software was altered.
Out of an abundance of caution we are taking steps to protect the process that certifies our macOS applications are legitimate OpenAI apps.
We are updating our security certifications, which will require all macOS users to update their OpenAI apps to the latest versions. This helps prevent any risk—however unlikely—of someone attempting to distribute a fake app that appears to be from OpenAI. You can update safely through an in-app update or at the official links below. 🧵
We should really drop this 'our side is purer' narrative. Pedophilia & abuse happens everywhere.
There are thousand cases with Hindu Culprits doing the exact same thing right here in this country without traveling to a some island.
There are Muslims in Epstein list
There are Jews in Epstein list
There are Christians in Epstein list.
But there is not a single Hindu in Epstein List.
We're Not Same.
97
This brought back some memories. My college partnered with IITB and was tasked with developing apps for Akash Tabs. We got some tabs for this. Most of them were broken or had backup of 20 mins. While the tablet was nothing, but a scam, IITB made a good course on app development.
Found this scam called Akash Tablet launched by Kapil Sibal when he was the Telecom Minister. I'd purchased it. Never worked properly even for a day.
5
345
Being a pet parent taught me one thing for sure that our general public is the most uninformed when it comes to pet care and pet behaviour understanding.
Hopefully someone takes this poor soul to vet.
📍Bijnor: A dog has been continuously performing parikrama of Hanumanji Murti for over 36 hours, resting briefly before resuming.
Devotees have gathered in large numbers to witness this devotional moment ❤️
Readers added context they thought people might want to know
Dog walking in circles is a sign of severe neurological disorder and needs immediate veterinary care.
sevneurology.com/blog/dog-walki…
283
India has a big problem of arm chair experts; they pretend to be expert and offer opinions on topic they have no Idea about.
@ShashiTharoor is the chair of Parliamentary Standing Committee on External Affairs.
He is required to attend state functions irrespective of his party.
Hello @ShashiTharoor
Now that u have crossed all limits by attending the banquet where your own party leader (LoP) wasn’t invited…….just wanted to ask
Are u dancing at Jindal’s wedding too??
153
Since consumer Internet is cheaper people think business Internet is similar l but that’s completely incorrect.
Most ISP refused to offer business broadband to large companies or offices so they have no option but buy a lease line which costs a minimum of 5LPA for 300 Mbps.
Replying to @shantanugoel
Bhai 1,000/month mein high speed fibre connection aa jata ab.
50 router laga lo usi connection mein
but they still throttle it as if we are using the costly Internet from 1998
45
You got a CNAPP, now what?
deepshankaryadav.net/you-got…
52
The real test for any self driving system would be India.
No matter how many variables you think of there will always be something new on the road.
Road being converted into Potholes suddenly.
People cutting into lanes without any indicator.
FSD is being tested with the water cup challenge. It's never dull in China. #FSDChina
Credit: Douyin 王船船
1
1
82
















