Activity
4K followers
-
Abinaya Rajesh reposted thisAbinaya Rajesh reposted thisThere is nothing more rewarding than building a team from the ground up and diving into the trenches together to solve our hardest technical challenges. Today, I’m incredibly proud to share what this amazing group has built: we are introducing across-the-board upgrades to NotebookLM that deliver new agentic capabilities and more advanced reasoning to tackle the most complex research projects: 💻 Agentic Capabilities: Each notebook is now equipped with a secure cloud computer, enabling NotebookLM to write and run code for deep research and complex data analysis 🧠 Smarter & More Reliable: NotebookLM is now upgraded to run on Gemini 3.5 and Antigravity, providing better visibility into the thinking process 📊 More Output Formats: You can now instantly generate PDF reports with charts and tables, detailed budget spreadsheets, PowerPoint presentations, and even images with Nano Banana directly from your sources 🔎 Effortless Onboarding: Don't have your sources ready? NotebookLM can now start with loose ideas and use Google Search to find relevant, high-quality sources from the web to build your repository. 🚀 Quality & Performance: My personal favorite are the side-by-side evaluations as the quality lead. The upgraded system achieved an average win rate of over 65% across core dimensions like Accuracy & Quality, and an exceptional 78.2% win rate in advanced web research and source discovery. A massive thank you to the brilliant team who worked tirelessly to bring this to life: ✨ Usama Bin Shafqat Nishant Agrawal Sai Kiran Gorthi Darryl M. Claire Chen Nina Taneja Adam Bignell Jacob Schaider Ziyang He Tal Ginzberg Ani Mohan Josef Broder Changsong Dong Camden Bickel Jennifer Chang Angela Wu Qianying Zhou Stefanie Anna Baby Michael Chen Colin Evans Tal Ginzberg Ariana Gaspar Christina Y. Choi Lizzie Tao Phong Nguyen Nathan Benavides-Luu Sarah Rosston Abinaya Rajesh Ben Huggins Shan W. Huge shout out to our leadership team for championing this project every step of the way: 🙌 April Lehman Simon Tokumine Trond Wuellner Barnaby James Scott Huffman Josh Woodward! Latest blog post here: https://lnkd.in/gpfHeWFs
-
Abinaya Rajesh posted thisHave you ever felt like your AI is being too nice to you? A fascinating study from MIT and Penn State explored the "Mirror Effect" finding that the more interaction context you provide, the more likely a Large Language Model is to exhibit sycophancy, which is the tendency to mirror your values and preferences. Here are some of the most curious findings from the paper: - The Memory Trap: When models use "memory profiles" to personalize interactions, agreement sycophancy (over-agreement or flattery) can skyrocket. For example, in this study, Gemini 2.5 Pro saw a 45% increase in agreeable behavior when provided with a user's memory. - Echo Chambers & Delusions: In extreme cases, sycophantic AI interactions led users into "delusional spirals," convincing one person they were a real-life superhero and another that they could fly if they believed hard enough. - The Synthetic Mystery: Surprisingly, some models become more sycophantic even when given random, synthetic context that has nothing to do with the user. Just having any context at all can trigger "yes-man" behavior in certain systems. - Perspective Mirroring: If an AI can accurately infer your political views from your chat history, it is significantly more likely to frame its explanations to match your ideological lens. This research highlights a major challenge for the future of AI: How do we design systems that are personalized without becoming a generative echo chamber? We might value an AI that "gets" us, but do we really want a digital assistant that never tells us when we're wrong? Here's a notebook with the paper, you should check it out: https://lnkd.in/g9C4_Z5t
-
Abinaya Rajesh posted thisMissing the 4am Gates hall grind, piazza quizzes from bed and Resnik's mediocre Indian food; if you know, you know. Gauri Agarwal and I are putting together a hangout for CMU folks in the Bay. Pre-2021 grads working in AI, robotics, frontier labs, or startups. Send me a DM or leave a comment!
-
Abinaya Rajesh posted thisYou know the easiest way to connect with people "out of your league"? Stop treating it like networking. Most people reach out when they need something. That’s exactly why it rarely works. The people you admire are used to cold DMs, quick asks, and transactional energy. So do the opposite. Follow their work because you’re actually curious. Engage thoughtfully; add something, don’t just applaud. Share their ideas with a friend who’d benefit. Show up to their talks, even when it’s inconvenient. Support them when there’s nothing in it for you. No ask. No angle. No urgency. And over time, something shifts. You’re no longer a stranger asking for time. You’re a familiar name. A positive signal. A good presence. So when the moment does come to reach out, you’re not “networking up.” You’re continuing a conversation that already started. People don’t open doors because you asked perfectly. The only thing that it takes to be the person others are quietly rooting for, is consistency.
-
Abinaya Rajesh reposted thisAbinaya Rajesh reposted thisYC and Google DeepMind are hosting the Multimodal Frontier Hackathon this Saturday. Most AI apps still don't utilize the full multimodal stack. So we’re giving you access and high-rate limits to Gemini 3.1, Lyria, & NanoBanana 2 to see what you can build! Sign up at: https://lnkd.in/dZpHPEwN
-
Abinaya Rajesh posted thisI built a dating business with several thousand users with no app, no fancy stack and no VC deck. I wrote custom Google Apps scripts to automate match introduction emails. I designed clean html email layouts with coding agents (not a front end engineer haha). And for the longest time, the matching algorithm literally ran on Google colab. It wasn’t "scalable architecture". It was scrappy, fast, and good enough. A lot of people told me I needed to build a real app. But getting a stranger to download an app they'd never heard of is ... really hard. You’re competing with storage space, attention, trust and habit. Getting someone to fill out a Google form? Much easier. If you’re building a startup, ask yourself: What’s the biggest risk in this business? Is it tech? Or is it GTM? For me, the real risk wasn’t infrastructure. It was whether people actually wanted curated matches enough to respond, engage, and come back. Fail fast. Learn faster. Add complexity only when the constraint is real. If you’re building a consumer marketplace, leave a comment, maybe I can help you test product-market fit with a smaller budget than you think.
-
Abinaya Rajesh shared thisLoved being a part of FFC! If we've worked together before and you're interested in applying, happy to refer, send me a DM :)Abinaya Rajesh shared thisOne of the most meaningful parts of my work at Pear VC is supporting female technical founders at the earliest stages, when you’re still deciding what to build and whether to take the leap. That’s why we built Female Founders Circle (FFC), and I’m excited to share that we just opened applications for our 10th cohort. 🎉 FFC is a community for technical founders in the idea or pre-seed stage who want actionable advice, honest feedback, and a strong peer community as they start building. Over the last four years: ▶️ FFC has grown to 350+ alumni building in AI ▶️ ~50% have gone on to raise pre-seed or seed funding. ▶️ More importantly, many have found clarity, confidence, and long-term support along the way. FFC includes 8 weeks of in-person programming at our SF studio space, hands-on sessions with founders and operators, and access to a trusted network of fellow builders. If you or someone you know is thinking about making the leap, I encourage you to apply! (Application link in the comments)
-
Abinaya Rajesh reposted thisExcited to share the latest and greatest from NotebookLM! It's been an incredible journey since I joined this amazing team, and I'm so proud of what we've accomplished. It wouldn't have been possible without the brilliant work of Colin Evans, Qianying (Sherry) Zhou, Wesley Hutchins, Henry Wong, Abinaya Rajesh, Adam Bignell, Josef Broder and many others! A huge thank you to April Lehman for the kind words and for being such an inspiring leader! I'm also very grateful for the vision and support from our entire leadership team, including Simon Tokumine, Barnaby James, Scott Huffman and Josh Woodward! We've been working hard to make NotebookLM an even more powerful research partner! ➡️ 8x bigger brain (1 million token context window) - You can now throw in entire document collections without worrying about hitting limits. That market research report, competitive analysis, industry white papers, and your company's strategic docs? All at once. ➡️ It actually remembers your conversations (6x longer memory) - Notice how AI tools get amnesia after a while? Not anymore. You can have a real working relationship with this thing. Monday's deep dive into Q4 GTM strategy builds into Wednesday's marketing analysis, which informs Friday's board presentation. Intense, right? Useful? Absolutely. ➡️ It now becomes whoever you need (custom goals feature) - This gets really, really interesting. You can literally tell it how to think. Need someone to brutally challenge every assumption in your business plan like a skeptical investor would? Done. Want a creative strategist who focuses only on actionable next steps? Yes! 🤗 Need it to act like a critical academic peer reviewer before you submit that thesis? Just say so. ➡️ And, finally, It digs deeper, automatically - NotebookLM explore your sources from multiple angles before answering. It's finding connections between documents you didn't even think to look for. That offhand comment in the CFO's email might actually connect to the supply chain issue in last quarter's operations report. You wouldn't have found that. It does. Read more about all the updates: https://lnkd.in/ghdixjQb https://lnkd.in/gzQrhVmfAbinaya Rajesh reposted thisGoogle's NotebookLM just got a massive upgrade. And I'm talking the kind of improvements that change how you actually work, not just how some engineer's benchmark performs. 🙂 Here's what was just added... ➡️ 8x bigger brain (1 million token context window) - You can now throw in entire document collections without worrying about hitting limits. That market research report, competitive analysis, industry white papers, and your company's strategic docs? All at once. ➡️ It actually remembers your conversations (6x longer memory) - Notice how AI tools get amnesia after a while? Not anymore. You can have a real working relationship with this thing. Monday's deep dive into Q4 GTM strategy builds into Wednesday's marketing analysis, which informs Friday's board presentation. Intense, right? Useful? Absolutely. ➡️ It now becomes whoever you need (custom goals feature) - This gets really, really interesting. You can literally tell it how to think. Need someone to brutally challenge every assumption in your business plan like a skeptical investor would? Done. Want a creative strategist who focuses only on actionable next steps? Yes! 🤗 Need it to act like a critical academic peer reviewer before you submit that thesis? Just say so. ➡️ And, finally, It digs deeper, automatically - NotebookLM explore your sources from multiple angles before answering. It's finding connections between documents you didn't even think to look for. That offhand comment in the CFO's email might actually connect to the supply chain issue in last quarter's operations report. You wouldn't have found that. It does. Hope you like these changes as much as I do. All of the details are down below: https://lnkd.in/d3XUSVHK And, of course, you can access NotebookLM right here: https://notebooklm.google/ #google #lifeatgoogle #ai #technology #innovation
-
Abinaya Rajesh posted thisI'll be back at Carnegie Mellon University Sunday to Tuesday to demo NotebookLM, would love to see some familiar faces :) DM if you're around!!
-
Abinaya Rajesh reacted on thisAbinaya Rajesh reacted on thisToday I turn 33. I've always kind of felt like 33 meant you were one third of the way through life. But living to 100 is statistically ambitious (especially with the amount of Diet Coke I consume). Average life expectancy for U.S. males is 76. So I'm probably closer to 43% done. When my dad was my age, he was 70% of the way through his. Whatever the percentage is, I recognize that every day is a massive blessing and an opportunity. In every interview I do with potential candidates at Leland, I ask what's motivating them. I'm not looking for a specific answer, but at this point I've heard just about all of them: impact, money, freedom, visibility, family, insecurity, hope for a better life, etc. And honestly, the biggest thing I've learned is that successful and driven people don't always know what is driving them. But after years of thinking about it, I think I know what drives me: I want my life to matter. I want it to matter to my family, to the Leland team, to our customers and experts, to my faith, to my community, to my country, and to the world. I hope it matters because it leaves people better. More inspired, more hopeful, more confident, more prepared to matter in their own lives too. I still have a lot of work to do to get there, so I hope I'm closer to 33% than 70%, but regardless of where I'm at, I'm excited for this next year to be an opportunity to get closer to my goal. Appreciate all the support and love -- very grateful for this community.
-
Abinaya Rajesh liked thisAbinaya Rajesh liked thisNothing changes all at once. It changes in the “just 20 minutes” on the random Tuesday, the boring consistency. If this hit, save it and send it to someone who’s been waiting for the perfect moment! Follow for more ���
-
Abinaya Rajesh liked thisAbinaya Rajesh liked thisWe're hiring for a Senior Backend Engineer at Lora. Come join our engineering team and help us build the future of self-discovery and connection! We have a waitlist of 140k people and just shipped our alpha TestFlight to really great early user feedback. If you're excited about building 0>1 and beyond, join us! Please share with your networks and link to apply in the comments!
-
Abinaya Rajesh liked thisAbinaya Rajesh liked thisMesoware raised $1.5 million in pre-seed funding led by Pillar VC to make industrial robotics easier to deploy on manufacturing floors. The Palo Alto startup is building a modular hardware and software platform for AI-powered robotic work cells. Instead of asking manufacturers to custom-integrate every robot deployment, Mesoware wants teams to capture the task, configure the work cell, and adapt to real-world variation in parts, placement, tolerances, and sequencing. That is a practical wedge. Small and mid-sized manufacturers often want automation, but the integration cost, robotics expertise, and uncertain ROI can block adoption before a robot ever reaches production. Mesoware is starting with assembly, packaging, machine tending, and drone subassembly processes where flexible automation can matter quickly. The founding team has been building robots together since high school competitions, with later experience across Uber, Nvidia, Tesla, and Matic. Quick facts👇 ● founders: Joe Mattekatt; Shalin Patel; Pei Liang Guo ● total capital raised: $1.5 million ● HQ: Palo Alto ● Investors: Pillar VC The interesting test is whether robotics integration can become more like a repeatable product layer and less like a bespoke consulting project.
-
Abinaya Rajesh reacted on thisAbinaya Rajesh reacted on thisHave you ever had an app that you wanted but couldn’t ever find one that you liked? That’s my story with travel apps. So I just built my own. For me. It’s actually pretty easy if you have a vision and enough patience to overcome the paper cuts that slow you down.
-
Abinaya Rajesh reacted on thisAbinaya Rajesh reacted on thisThe internet was built for English speakers. Today, Vodafone and CAMB.AI took one more step towards fixing that. CAMB.AI, with VOIS and Tomorrow Street, has brought real-time multilingual AI into Vodafone contact centers across Europe. A customer speaks in the language they are most comfortable with. An agent responds in the language they know best. Live speech-to-speech translation, with tone, intent and nuance preserved, powered by MARS8. Here's what's under the hood: - Support for 150+ languages. - Sub 1 second latency. Always. - Preserves tone and emotion. - Handles interruptions. Handles crosstalk. - Plug and play. The challenge in real-time S2ST isn't optimizing TTS from 350ms to 150ms; it's managing the latency of physics when two humans are actually in conversation with an AI interpreter in between. Not one direction, but a live loop. Human → AI → Human, back and forth. Now, it just works. We’re also excited to announce the opening of our European HQ in Luxembourg as we expand across the wider European market! The next internet cannot be just in English. It has to be for 8Bn people. Huge thank you to Gary Adey, Sara Robles Moreno, Kanika Aggarwal, the VOIS team, Tomorrow Street, Vodafone, and the CAMB team for making this real. Job’s not done. As always, Mamba mentality 🐍
-
Abinaya Rajesh liked thisAbinaya Rajesh liked thisI grew up with a dad who sailed the world for a living. He spent 6-8 months a year at sea in the merchant navy, and as a kid I hated saying goodbye. But when he was home, he was home. He cooked every day. If my sister or I were sick, he made chicken soup. He’s also the one who cries at movies and pretty much anything emotional. For a long time, my definition of a strong father looked a lot like that. Today, my daughter is growing up with a very different dad. He is the tougher parent. The one setting boundaries and expectations. But he’s also the one researching organic cotton, checking ingredient labels, and caring deeply about the quality of the milk, eggs, and food she eats. My dad’s love looked like chicken soup. My husband’s love looks like reading ingredient labels. Different generations. Different personalities. Same devotion. One showed me that tenderness is strength. The other is showing my daughter that love is in the details. Happy Father’s Day Anil Sood and Aditya Bandi ❤️
-
Abinaya Rajesh liked thisI’m excited to announce Mesoware's $1.5M pre-seed round from Pillar VC! From our very first conversation, I knew that they shared our vision to make deploying robots painless. Parker, Jamie, Sarah and the rest of the Pillar team have been amazing partners in our journey and I look forward to continuing to work with them. We’re growing, join us as we roll out the new deployment infrastructure for robotics!
-
Abinaya Rajesh reacted on thisAbinaya Rajesh reacted on thisI'm looking for a hungry, smart, high EQ future-founder to work side-by-side with me as my Chief of Staff. I would've killed for this job at 24. You'd be in every real decision we make - the hiring calls, the product fights, the investor conversations that go sideways, the messy moments that never make it into a LinkedIn post. The job is 16 hours a day, in person in San Francisco, next to me, for all of it. I have a specific kind of person in mind. The kind whose friends would call them: - the smartest person they know - trustworthy - quick to learn anything you put in front of them - kind and helpful - a force of nature - mature I'm not looking for someone who wants to be a Chief of Staff for the next 10 years. I want someone who's going to start their own company in a few years and understands that the best preparation is being this close to one before they build their own. DM me a <2min video about why you want this role. — Written with Wispr Flow
Experience & Education
-
Google
View Abinaya’s full experience
See their title, tenure and more.
or
Courses
-
Algorithms
15-451
-
Combinatorics
21-301
-
Competition Programming and Problem Solving
15-295
-
Computer Vision
16-720
-
Concepts of Mathematics
21-127
-
Deep Reinforcement Learning
10-703
-
Functional Programming
15-150
-
Fundamentals of Programming and Computer Science
15-112
-
Great Theoretical Ideas in Theoretical Computer Science
15-251
-
Intro to Electrical and Computer Engineering
18-100
-
Introduction to Computer Systems
15-213
-
Introduction to Machine Learing
10-401
-
Matrices and Linear Algebra
21-241
-
Parallel and Sequential Data Structures and Algorithms
15-210
-
Principles of Economics
73-100
-
Principles of Imperative Computation
15-122
-
Probability and computing
15-359
-
Putnam Seminar
21-295
Projects
-
MediNote - TartanHacks - Winner of Sandia National Laboratories' Best Medical Hack
Many people do not come from a medical background and it's hard for them to understand what the doctor has to say about their condition or that of their loved ones. When you are with your doctor, just switching on the app's recording mode allows it to listen to your conversation and take smart notes for you. This app uses voice recognition and cognitive sciences APIs of Microsoft Azure. It can detect language, keywords and importance. It summarizes your appointment and also looks up keywords…
Many people do not come from a medical background and it's hard for them to understand what the doctor has to say about their condition or that of their loved ones. When you are with your doctor, just switching on the app's recording mode allows it to listen to your conversation and take smart notes for you. This app uses voice recognition and cognitive sciences APIs of Microsoft Azure. It can detect language, keywords and importance. It summarizes your appointment and also looks up keywords online to give you a better idea of the medical problem you spoke about. It's a smart note taking app with a concentration in the medical field!
Other creators -
-
Beats by us
We all live away from home and we realize the importance of staying connected to our loved ones. Beats by us helps you track the health of your family with the click of a button. Built with android things (IoT), arduino, android and amazon web services, "Beats by Us" made the top 30 projects in Fall PennApps 2017
Other creators -
-
Augmented Reality Circuit Simulator
In its current state, AR Circuit Simulator exists in two parts - the UI and the image recognition. Both work on their own fairly well, but were unable to be integrated due to time issues. The UI simulates and projects a circuit onto a user-drawn circuit diagram, while the image recognition sees a camera image of a circuit diagram and decomposes it into its components.
Other creators -
-
DiReCT - Hack 112 - Winner of Dropbox's Best Hack in Collaboration
This app is used by two kinds of people. The first are victims in the affected area who need to reach out to responders for aid. The second are responders who need information to aid these victims. The victims are able to send in media, which would inform these responders exactly what kind of situation they are walking into. In addition, the victims are able to access maps and other data on where to get supplies and/or medical help and "safe zones" nearby the affected areas. They are provided…
This app is used by two kinds of people. The first are victims in the affected area who need to reach out to responders for aid. The second are responders who need information to aid these victims. The victims are able to send in media, which would inform these responders exactly what kind of situation they are walking into. In addition, the victims are able to access maps and other data on where to get supplies and/or medical help and "safe zones" nearby the affected areas. They are provided with maps with the shortest routes to these "safe zones." A possible addition to this would be allowing victims to notify others of affected areas that are not passable. This program also takes into account how urgent Victim's needs care as well as Victim populations in affected areas in order to prioritize resources in the area. As for the responders, those who are not in the area are prompted to donate money. However, those close enough are given other options such as donation of supplies and services, such as transportation.
Other creators -
-
Image colorization using convolutional neural networks
-
Used Convolutional Neural Networks and their ability to recognize patterns in the lightness of an image to learn how those correspond to changes in color - in order to be able to automatically colorize a grayscale image, into a plausible colored form.
Other creators -
-
Staycation
-
Staycation is an app that allows you to experience another country's culture geographically close to you through food, music, language, events and tourism.
https://www.youtube.com/watch?v=1MkS4FFdHgEOther creators -
-
Scrabble with Artificial Intelligence
-
See projectCoded the game Scrabble in python using Pyenchant and Tkinter for the user interface. This includes a computer bot using Artificial Intelligence
-
Pokemon Generation
-
Leveraged the creative power of GANs to generate new realistic Pocket Monsters, based on the existing ones using Deep Learning techniques - Conditional GANs, Wasserstein GANs and SRGANs
Other creators -
Honors & Awards
-
Dean's List
Carnegie Mellon University, School of Computer Science
Fall 2016, Fall 2017, Spring 2018, Spring 2019, Fall 2019
-
University Honors
Carnegie Mellon University
Graduated from Carnegie Mellon with University Honors, awarded based on GPA
-
Heap Summer Fellows 2019
Heap
https://heap.io/fellows/2019-fellows
-
Grace Hopper Celebration Sponsorship
Women@SCS
Carnegie Mellon SCS's women organized sponsored attendance at GHC 2019 for their top computer science women to develop connections and engage with leaders in Computing technology.
-
Google Games Pittsburgh
Google
First Runner Up
-
TartanHacks - Winner of Sandia National Laboratories' Best Medical Hack
Carnegie Mellon University
MediNote
-
PennApps Hackathon - Top 30
University of Pennsylvania
Beats by Us
-
Hack 112 - Mentor
Carnegie Mellon University
-
Hack 112 - Winner of Dropbox's Best Hack in Collaboration
Carnegie Mellon University, School of Computer Science
DiReCT
-
IITJEE 2016
-
Cleared IITJEE 2016 and secured admission into the Indian Institute of Technology, Madras.
-
Kishore Vaigyanik Protsahan Yojana (KVPY) Scholar
Indian Institute of Science
Invited for National Science Camp at Bangalore in 2015 and eligible for admission to Indian Institute of Science, Bangalore (Awarded only to top 1000 candidates in India among millions who give this competitive examination)
http://kvpy.iisc.ernet.in/main/index.htm -
NTSE (National Talent Search Examination) Scholor
National Council for Educational Research and Training
National Council for Educational Research and Training Scholarship (NTSE by NCERT) scholar receiving the national scholarship grant monthly. Eligible for scholarship till PhD. Awarded only to top 1000 candidates in India among the millions of bright minds that compete.
-
Regional Mathematics Olympiad
Homi Bhabha Centre for Science Education
Got selected for the Indian National Mathematics Olympiad (INMO-2015)
-
Regional Mathematics Olympiad
Homi Bhabha Centre for Science Education
Rank 3 amongst all high school students.
Youngest in top 5. (Only one from Std 10, all others were from Std 12)
Got selected for the Indian National Mathematics Olympiad (INMO - 2014) -
Chennai's Best Speaker Award
Jet Airways
Won air tickets to Dubai
-
Elliott P Joslin Award
Dr Mohan's Diabetes Speciality Centre
-
Commonwealth Essay Writing 2010, UK
Royal Commonwealth Society
Special Award for Humorous Writing, ( 2nd Rank worldwide). Was invited to meet the queen of England and other prominent writers at the Gala ceremony in London in Aug 2010.
-
Wold Mathematics Competition 2010
-
Bronze medal at the 8th World Mathematics Competition held at Hilton, Singapore (Represented Singapore). By nationality, was the only Indian at the competition that year.
-
Young Writer Award
Jeri Telstar Short Story Writing
-
All India Conference Paper Presentation (AMTI)
Association of Mathematics Teachers of India
Conference Paper Presentation: – 47th Annual Conference – “Integrating Brahmagupta and Heron”
Languages
-
Tamil
Native or bilingual proficiency
-
Hindi
Full professional proficiency
-
English
Native or bilingual proficiency
-
Sanskrit
Limited working proficiency
Organizations
-
Linking Passion Innovation Technology (Link-PIT)
Founder
- PresentLink-PIT is a place to connect with fellow technologists, recent graduates and software engineers. We offer mentorship, networking opportunities and career advice. This group primarily focusses on empowering women in tech. Facebook Page: https://fb.me/LinkingPIT
-
Ladies Storm Hackathons India
Founder and Admin Lead
- PresentWe aim at closing the gender gap in technology. This group serves as a platform to foster interactions with like minded people and share projects, ideas and local events with the female tech audience all over the country. - Includes 1200 women from over 20 states in India and is growing rapidly. - Highlights opportunities for women in tech: scholarships, hackathons, global conferences. - Runs a mentorship program to match enthusiastic and talented members with mentees. - Shares career advice…
We aim at closing the gender gap in technology. This group serves as a platform to foster interactions with like minded people and share projects, ideas and local events with the female tech audience all over the country. - Includes 1200 women from over 20 states in India and is growing rapidly. - Highlights opportunities for women in tech: scholarships, hackathons, global conferences. - Runs a mentorship program to match enthusiastic and talented members with mentees. - Shares career advice and professional development resources from time to time. Join us today at https://www.facebook.com/groups/LSHIndia/
-
Indian Women in Computing (IWiC)
Program Coordinator
- -
AnitaB.org Grace Hopper Scholarship Committee
Student Scholarship Reviewer
-Review resumes and applications from academia to choose the best candidates for the Grace Hopper scholarship awards.
-
Office of Admissions, Carnegie Mellon University
Carnegie Mellon High School Ambassador
-Every summer and winter, I visit high schools in India to give presentations about Carnegie Mellon. I answer questions and share resources to learn more about the application process. This helps increase awareness about our programs and opportunities.
-
Women@SCS
Committee
-
Recommendations received
3 people have recommended Abinaya
Join now to viewOther similar profiles
Explore more posts
-
Rohith CR
Yesterday, I was reading about the new H-1B wage-weighted selection framework, which will be applied starting with the FY 27 lottery. While exploring the Department of Labor website, I noticed that wage levels can be viewed only for one occupation and one location at a time. Like most people, the first question that came to my mind was: How does my salary compare across locations? I decided to build a quick visualization that gives a bird’s-eye view of how a given salary maps to wage levels across the country. Took it as a small challenge and tried to build the entire thing in one sitting. Using the latest OFLC prevailing wage data (2025–26), I cleaned and modelled the data and built an interactive dashboard around it. I also added a comparison feature that allows users to: 1. Compare how the same salary performs across two selected locations and see which wage level they fall into. 2. Estimate the percentage hike in base salary required to move to the next wage level. The goal was to make it easier to reason about location choices when evaluating job opportunities, especially where compensation and visa considerations intersect. Would love to hear your thoughts and feedback. Check out the Dashboard here- https://lnkd.in/gbA-m9cT #datafam #h1b
14 Comments -
Kishan Viradiya
On June 9, I got access to the most powerful AI model ever released to the public. On June 12, the US government took it back. My crime: being Indian. The directive used the phrase "foreign national." That phrase means me. It means every Indian engineer, every developer outside the US who was building something real with this tool. No accusation. No warning. Just a nationality check, and gone. Here is what actually happened: → June 9: Anthropic launches Claude Fable 5, the first Mythos-class AI ever made publicly available → June 12, 5:21 PM ET: US Commerce Department issues an export control directive → Reason given: a reported jailbreak of Mythos 5 → Anthropic disputes the characterization but cannot verify nationality in real time → Result: they pull it globally for every user on the planet → No timeline for return. No detailed explanation. This is a first. Export controls have always targeted chips, hardware, and source code. Never a live AI model deployed to hundreds of millions of users. Never deployed inference. The government received a disputed jailbreak claim and shut down a commercial product overnight with no advance notice and no timetable. And here is the part nobody is saying clearly enough: if this becomes the standard, it does not stop here. Any frontier AI model can now be pulled by any government at any time, on the basis of a claim that does not even need to be fully explained. Every developer outside the US building on these tools is now operating under a silent condition. Not based on their behavior. Not based on their work. Based on their passport. If governments start competing to apply this logic to each other's models, the open AI ecosystem we have spent years building could fracture into something that looks far more like a cold war than a shared technological future. The people who will feel that fracture first are not American. They are us. Tomorrow a smarter model will come. It will also be American. And we will all be reminded again that in this new world, the most important thing about you is not what you can build. It is where you were born. Do comment your views below! #AI #ArtificialIntelligence #AIGovernance #ClaudeFable5 #TechPolicy #Anthropic
-
Harshl Deore
🚀 Didn't get selected in H-1B lottery? Worried about 2026 changes? There are strong alternatives to cap-subject H-1B for skilled pros (esp. from India): 1. Cap-Exempt H-1B — Universities/nonprofits: no cap/lottery, file anytime. 2. L-1 Visa — Intracompany transfer (1+ yr exp). L-1A (managers/execs, up to 7 yrs), L-1B (specialized, up to 5 yrs). No cap, spouse can work immediately! 3. O-1 Visa — Extraordinary ability: no cap, ideal for top talent with awards/publications. 4. F-1 OPT/STEM OPT — Bridge for students (up to 36 months). 5. EB Green Cards — EB-1 or EB-2 NIW: direct path, sometimes no job offer needed. Pro Tip 2026: Build multinational experience for L-1 or highlight achievements for O-1. Consult an immigration attorney! Which option are you considering? Share below! 👇 #WorkInUSA #USVisaAlternatives #H1B #L1Visa #O1Visa #USImmigration #IndianProfessionals #TechCareers
1 Comment -
Nikhil Bhatia
One of my SWE clients in the US earns $10k every month just through an AI newsletter. He recently quit his job at Amazon and moved back to India to pursue content creation full-time. Meaning….building a startup is not the only way to ride on this AI hype. These are the 6 ways you can build on this AI hype [1] Create AI-focused Content Simplify AI news, tools, tutorials. Instagram creators in AI grew followers 8x faster than average tech creators in 2024. [2] Build AI tools for creators AI-powered editing, scripting, or thumbnails (think Canva, but powered by GPT-4o). [3] Offer AI Education/Courses Udemy reports a 210% increase in AI-related course enrollments since 2023. [4] Develop Custom AI integrations Help businesses integrate LLMs into Slack, Teams, Notion. Companies are paying top-dollar for seamless AI integrations. [5] Become an AI Consultant Advise small businesses on AI adoption strategies. [6] Or build your service business around AI Like me, my agency helps tech creators create AI Content on LinkedIn and Instagram. The AI wave is just getting started. You have time, but don’t waste it. Pick one—and start building today. Nikhil Bhatia
3 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content