Anoop (@hashedDNA) header
Anoop (@hashedDNA) profile picture

non-populist views, minimalist, tech savvy, loves automobile, instant traveler, photographer, mission: better humanity.

Bangalore
Joined January 2009
Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)Photo by Anoop (@hashedDNA)
An IIT Madras-incubated startup is using AI to figure out how to manufacture a product, how much it should cost, and how to make it cheaper! Imagine designing a machine component on your computer. The design looks perfect. But how do you manufacture it? Should you use 3D printing, casting, forging or traditional machining? Which material would be most suitable? And how much would the entire process cost? Answering these questions can require considerable engineering expertise, time and money. That's precisely the problem Matterize, a Chennai-based deep-tech startup incubated at IIT Madras, is trying to solve. Founded in 2024, the company has emerged from IIT Madras's Nirmaan pre-incubation programme, with IIT Madras graduate Sarthak Saurav among its founders. Its solution is an AI-powered manufacturing intelligence platform. Here's how it works. An engineer uploads a 3D CAD model of a component onto the platform. Matterize analyses its geometry, evaluates different manufacturing methods and recommends suitable production processes. It then generates a detailed cost estimate, breaking down expenses such as raw materials, machine time, labour, tooling and overheads. The company says its platform can evaluate more than 50 manufacturing options! It can also identify design features that might make production unnecessarily expensive and estimate the carbon emissions associated with manufacturing. For companies developing complex products, such insights could be valuable. Engineers could compare manufacturing alternatives before production begins. Procurement teams could evaluate supplier quotations more effectively. Startups could identify costly design decisions before investing in expensive tooling. Matterize's emergence also reflects IIT Madras's growing deep-tech ecosystem. In April 2025, the institute announced that its incubation ecosystem had supported 104 new deep-tech startups during the preceding financial year, with Matterize among them. As India works towards becoming a global manufacturing powerhouse, technologies like these could help manufacturers make faster, more informed production decisions. Because sometimes, the biggest manufacturing savings begin long before a machine is switched on.
13
62
3
308
13,096
This song hits different Song : "Safal Hogi Teri Aradhana" Singer : S.D. Burman Movie : Aradhana (1969) Tribute to the legend SD Burman on his 51st death anniversary! 🙏 #SDBurman
26
115
7
296
35,309
At 20, Renee Noronha has taken on one of the toughest tests of endurance, completing the IRONMAN World Championship in Kona, Hawaii. On 10 October 2026, she made history as the youngest Indian woman to finish the gruelling race, which demands far more than physical strength. The challenge involved swimming 3.8 km in open water, cycling 180 km and then running a full 42.2 km marathon. That’s 226.2 km in total, through choppy seas, strong winds, rain and scorching heat. But Renee’s journey into endurance sports began with gymnastics. From the age of six to 16, she trained in the sport before embarking on a new athletic journey that would take her to the world stage. And she isn’t stopping here. Renee now aims to complete IRONMAN races across all six continents and pursue a Guinness World Records title. Credits : renee.noronha | ironmantri | ironmanasia on IG #ReneeNoronha #IRONMAN #SMEDIT [youngest Indian woman Ironman finisher, Renee Noronha Ironman World Championship, youngest Indian female Ironman Kona Hawaii, Indian women in Ironman triathlon, 20 year old Indian Ironman finisher]
12
116
5
603
16,491
42 years later, a new record! 🇮🇳⏱️ Vithya Ramraj has broken PT Usha’s long-standing national record in the women’s 400m hurdles, clocking 54.75s at #AichiNagoya2026. 🔥
23
4
119
8,376
Flipkart takes advertising to a new place. This time, inside Bengaluru's metro station with a projection and it moves like a train.
51
89
28
2,521
683,442
A RECORD-BREAKING VICTORY! 👏 Ladakh's own, Nawang Tsering wins the gruelling 122km-long Silk Route Ultra in 12:12:18 💥 An absolutely sensational performance!
33
374
85
2,334
446,536
The OECD conducts an assessment called PISA that measures the reading, science, and maths skills of 15-year-olds. The latest results are quite dismal. Across much of the world, kids are getting worse at reading and maths, and the decline isn't small. COVID probably explains some of this, but there seems to be a longer-term trend at play. And now there's AI. I can see a lot of this in myself. Pretty much every post I write now goes through ChatGPT, Claude, etc. Because I know AI will fix my grammar and typos, there's less incentive to get them right in the first place. And over time, I genuinely feel like my writing without AI is getting worse. I keep asking K (CTO) about coding and whether something similar is happening with our tech team. His answer isn't very reassuring either. 😬 All of this makes me wonder what the right approach is with kids. On the one hand, you don't want them outsourcing thinking, writing, problem-solving, etc., before they've learnt how to do these things themselves by making mistakes. On the other hand, if they're growing up in a world where AI will be a big part of their lives, you want them to be good at using it. The problem with humans is that, given a choice between an easy thing and a hard thing, we'll almost always choose the easy path. The same applies to AI. Even if it can help us think and do more, I think most of us will end up outsourcing a lot of our thinking and doing to it. I don't know how many here have watched Idiocracy, but sometimes it feels like we are accelerating towards it. 😬 PS: Idiocracy is about an average man who wakes up 500 years in the future to discover that human intelligence has declined so much that he is now the smartest person alive.🙂 Chart: @TheEconomist
60
81
18
514
50,156
I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.
20,394
166,706
39,464
799,165
175,642,178
Anoop retweeted
35 years strong. 🐧🧡 Happy birthday, Linux.
335
3,550
240
23,745
542,019
Anoop retweeted
About 1 in every 3 tonnes of fruit, vegetables and milk produced in India never reaches anybody. It rots on the way, because there was no cold storage and no refrigerated truck. The trucks that do have a fridge run it off the engine. So the engine has to keep running, burning diesel, even when the truck is standing still in traffic. A 25 year old engineer decided to take the fridge off the engine. Her name is Shaivee Malik. She studied biotechnology at the Indian Institute of Technology in Delhi. While she was there she joined the college formula racing team, Axlr8r, and became associate team captain. That is where she learned to build things that actually work, with 2 friends from the same team, Vivek Mahindrakar and Dharmik Bapodara. The 3 of them wanted to work on climate change. They looked around for where the damage was worst and least talked about, and landed on the cold chain. In 2022 they founded a company called Yotuh Energy. The name comes from Sanskrit and refers to the act of cleaning. They registered it in Delhi and then moved to Chennai for an incubation programme at IIT Madras. Their product is a refrigeration unit that runs on its own lithium ion battery, completely independent of the truck engine. That one change does several things at once. The fridge runs whether the engine is on or off, so the truck stops burning diesel while it idles. It keeps working even if the vehicle breaks down. Sensors on the doors detect every time a door opens and adjust the cooling for it. And because it is a separate unit, it can be fitted to an ordinary diesel vehicle or an electric one. It runs 6 to 8 hours on a charge. It cuts refrigeration running costs by around 75%. Each unit avoids somewhere between 3 and 36 tonnes of carbon dioxide a year, depending on use. They did not start with money. She has said the first support came from family. After that it was small government grants, 30,000 rupees a month for a year under one scheme, then a larger one, which is what paid for the early prototypes. Then came a funding round of about 1.53 crore rupees, and grants from the agriculture ministry, the science and technology department and others. Their paying customers now include distributors for Amul ice cream and for McCain frozen foods. She is clear about what the company actually is. We are a cooling solutions company, she says. Trucks are just where they started, because a fridge tied to a diesel engine is the least sensible way to keep something cold.
96
510
42
2,097
159,711
Proud to see our incubated deep-tech startup, The ePlane Company, pull back the curtain today on the e200X (PT-01), the full-scale physical reality of their physics-first approach to aviation. While the global industry has spent billions building massive aircraft that require cities to rebuild their infrastructure completely, IIT Madras faculty member Prof. Satya Chakravarthy and his team took a completely different path. At just 8m × 11m, the e200X is the world’s most compact passenger eVTOL, engineered to carry a pilot + 200kg of payload directly out of current helipads. Backed by over 10,000 km of cumulative flight testing and holding a formal Design Organisation Approval (DOA) from the DGCA, ePlane represents the elite tier of execution rigour in the Advanced Air Mobility sector. @iitmadras @ePlaneCompany @satchakra_iitm #IITMadras #IITMIncubationCell #DeepTech #MakeInIndia #eVTOL #AdvancedAirMobility #AerospaceEngineering #Innovation
32
305
19
1,438
47,347
Ashok Elluswamy, Tesla’s VP of AI, confirms the team is actively working on letting you prompt FSD with natural language. That’s gonna be so epic. "Pull into the driveway and park behind the white car.” So many possibilities.
121
372
26
5,059
352,071
For those tracking the America closed source vs China open weight model debate: In an initial pilot on modernizing an application from PHP to Next.js, Opus 4.8 with 8090’s Software Factory was simultaneously 1.4× cheaper and 1.5× faster than Opus 4.8 alone. Pairing our Software Factory with the cheaper GLM 5.2 model cut costs 16.4×, though it ran 3× slower than Opus 4.8 alone. Results are directional (n=1 per arm), and next steps are to rerun with proper controls and 10–15 buyer-relevant legacy modernization tasks developed with our Sales and GTM teams. More to come but the important question it raises is why any American public company with shareholders burning money on closed source models when the open weight ones are so much cheaper? We will rerun this on American open source models next (aka Nvidia)…
158
121
29
1,836
676,849
Video footage of the exact moment the BYD Yangwang u9 Xtreme reached a maximum speed of 496.22 km/h on the Track in Petersburg,Germany, officially beating the Bugatti Chiron super sport’s record as the world’s fastest production car
752
6,316
935
42,533
5,291,662
Repairing Underground Cables. NO Road Damage At numerous locations within Brookefields Layout, we have installed ducts and chambers to access underground cables and valves. Last Monday was the most recent cable fault. Where typically the road is hacked out and a crater left behind, here we simply un-weld the duct lid. When the underground cable is mended, we re-weld the lid. No road damage. Minimal traffic stoppage. Aside from allowing access, the duct, made of structural steel, protects 8-10 cable joints from hammering by ongoing traffic, and has reduced cable faults from once per month to once a year. A reminder that this duct was built with locally sourced money (Rs. 10 lakhs), and its design was provided by a local structural engineer. Local engineering talent. Local funding. Local execution. Local problem-solving.
4
14
4
70
14,265
Rename any .exe to .pdf right now. Email it to yourself. Your email client will probably show a PDF icon. That file extension is a label anyone can type. Your operating system still trusts it. Attackers know this. Google processes hundreds of billions of files per week across Gmail, Drive, and Safe Browsing. They built a tool years ago to see through fake extensions. It's called Magika, they open-sourced it, and the way it works is worth understanding. For 50 years, the standard for detecting what a file actually is has been the Unix `file` command. It reads "magic bytes," the first few bytes that act as a format fingerprint. PDFs start with `%PDF`. JPEGs start with `FF D8 FF`. ZIPs start with `PK`. Attackers figured out how to spoof these decades ago. A polyglot file can start with valid JPEG magic bytes but embed executable code deeper in. The header check passes. The antivirus waves it through. The payload executes. Magika takes a completely different approach. A 1MB deep learning model trained on 100 million files across 200+ content types. Instead of checking a header, it reads the structural patterns of the content itself. The way a forensic examiner identifies a document by its internal structure. 99% accuracy. 5ms per file on a single CPU. Near-constant inference time regardless of file size, because the model only reads a small subset of the content. A 5KB config file and a 5GB archive take the same time to classify. The biggest gain over traditional tools: textual file types. Source code, config files, markup. Traditional detection misses roughly 20% of these. Magika cut that gap almost entirely. Those are exactly the file types that supply chain attacks hide inside. Shipped in Rust. Integrated with VirusTotal. Apache 2.0. One command: `pip install magika`. The `file` command is 53 years old. This is the first real replacement.
11
74
4
603
81,379
Today, India takes a defining step in its civil nuclear journey, advancing the second stage of its nuclear programme. The indigenously designed and built Prototype Fast Breeder Reactor at Kalpakkam has attained criticality. This advanced reactor, capable of producing more fuel than it consumes, reflects the depth of our scientific capability and the strength of our engineering enterprise. It is a decisive step towards harnessing our vast thorium reserves in the third stage of the programme. A proud moment for India. Congratulations to our scientists and engineers.
4,453
28,742
3,206
171,351
41,710,837
A post reposted by Japan’s Finance Minister Katayama. Japan’s major oil company ENEOS has succeeded for the first time in producing synthetic fuel (e-fuel) — made without using petroleum — from CO₂ in the air and water at its facility in Yokohama. The demonstration plant’s current production scale is still very small at 1 barrel per day (roughly the size of one drum), but the company aims to scale it up to 10,000 barrels per day by 2040. Although it is still at the experimental stage, this is drawing attention as a new fuel that could reduce dependence on petroleum. In the timeline we’re currently living in, the move away from oil appears to be beginning at the same time as the move away from the US dollar.
合成燃料は、故渡文明元石連会長・エネオス会長の、水素と並んでの肝入りでもあられたと。私の衆議院議員時代の後援会代表世話人で、自民党の総合エネルギー調査会でもお話しお聞かせ頂きました。未来への投資!
389
2,845
371
14,320
1,927,834
Anoop retweeted
the generation that refused to accept cookies. is now giving AI access to their desktops, files, and bank accounts.
339
1,863
166
14,447
328,656