Lear Labs reposted this
We're looking to back founders and researchers building at the intersection of AI × Bio in India 🇮🇳 If you're building something interesting or know someone who is, please reach out or leave a comment. We'd love to fund you.
We build vertical AI products with the context of the critical workflows, domain expertise, specialized instructions, and data of each industry. We build anticipating what's possible with AI models in a year from now, not just today.
External link for Lear Labs
Workflo Ranka Junction
Bengaluru, IN
Lear Labs reposted this
On Friday, our CEO Zaid S. took the stage at National Law Institute University, Bhopal for the NLIU - Justice R.K. Tankha Memorial International Arbitration Moot 2026. Speaking to a room full of law students from across the country, he shared insights on the future of legal work in the age of AI, highlighting how the legal market is set to expand, and how this growth will unlock a wide range of opportunities for the next generation of lawyers across diverse domains. We were also excited to announce the launch of campus.atherna.ai — a free student version of Atherna, now open to campuses nationwide. We’re looking forward to seeing how students leverage AI to learn faster, think sharper, and build the future of law. Atherna was proud to serve as the Technology Sponsor for the competition and will continue to support and empower the next generation of lawyers across the country. #LegalAI #FutureOfLaw #Atherna #NLIU #LegalTech
Lear Labs reposted this
The 2,000-page document that started Atherna February 4, 2025, exactly an year ago I got a call from an old friend, a partner at a top-tier Indian law firm. They were preparing for a major arbitration, and the case record had crossed 2,000 pages - pleadings, annexures, witness statements, past judgments. They wanted to see if AI could help them cut some workload and get to the facts faster. Every AI tool they tried either: Refused to upload the file, Silently ignored parts of it, Or asked them to “split the document into smaller parts.” The model doesn’t say “I don’t know.” It answers anyway. It synthesizes partial truth into a coherent lie. And because the output sounds polished, the danger goes unnoticed. Arguments get built on omissions. When you split a case record, you don’t simplify it. You amputate it. A concession on page 743 no longer qualifies an argument on page 7. A witness statement loses the annexure that neutralizes it. Precedent is cited without the exception that defeats it. “The AI lost the context. It didn't know P-4 was connected to the affidavit.” That was not the only problem. Most of the record was scanned. Not searchable PDFs. Scanned copies of filings, orders, photocopied annexures, documents scanned years ago, sometimes scanned multiple times over. This isn’t an edge case. This is the Indian legal reality. Our courts and tribunals still run on scanned records. Context lives across thousands of faded, misaligned, non-OCR pages, not clean text files or neatly structured data. “If the system can’t read the entire record,” my friend said, “it shouldn’t give an answer at all.” I agreed. That conversation is what started Atherna, a legal AI built for Indian reality, designed to preserve context, surface uncertainty, and earn trust instead of faking certainty. If you are an Indian lawyer dealing with limitations with existing AI tools, feel free to reach out. We are building Atherna for you.
Lear Labs reposted this
A very interesting story about how we started Atherna. I also remember a senior lawyer telling us early on: “We don’t need a work product that looks right, we need one that is right.” That line has stayed with us, and it has quietly shaped how we’ve approached building this product.
AI | Machine Learning | Science & Tech Communicator With 1M+ Followers | Co-founder & CTO @ Atherna - an AI Co-counsel for Lawyers
The 2,000-page document that started Atherna February 4, 2025, exactly an year ago I got a call from an old friend, a partner at a top-tier Indian law firm. They were preparing for a major arbitration, and the case record had crossed 2,000 pages - pleadings, annexures, witness statements, past judgments. They wanted to see if AI could help them cut some workload and get to the facts faster. Every AI tool they tried either: Refused to upload the file, Silently ignored parts of it, Or asked them to “split the document into smaller parts.” The model doesn’t say “I don’t know.” It answers anyway. It synthesizes partial truth into a coherent lie. And because the output sounds polished, the danger goes unnoticed. Arguments get built on omissions. When you split a case record, you don’t simplify it. You amputate it. A concession on page 743 no longer qualifies an argument on page 7. A witness statement loses the annexure that neutralizes it. Precedent is cited without the exception that defeats it. “The AI lost the context. It didn't know P-4 was connected to the affidavit.” That was not the only problem. Most of the record was scanned. Not searchable PDFs. Scanned copies of filings, orders, photocopied annexures, documents scanned years ago, sometimes scanned multiple times over. This isn’t an edge case. This is the Indian legal reality. Our courts and tribunals still run on scanned records. Context lives across thousands of faded, misaligned, non-OCR pages, not clean text files or neatly structured data. “If the system can’t read the entire record,” my friend said, “it shouldn’t give an answer at all.” I agreed. That conversation is what started Atherna, a legal AI built for Indian reality, designed to preserve context, surface uncertainty, and earn trust instead of faking certainty. If you are an Indian lawyer dealing with limitations with existing AI tools, feel free to reach out. We are building Atherna for you.
We're hiring a growth-savvy intern for our portfolio company Atherna — the AI co-pilot for lawyers. If you enjoy experimenting, running scrappy campaigns, and using AI to push your output 10×, you’ll love this role.
We’re looking for a sharp, creative and ridiculously hands-on Growth Intern for Atherna. If you are someone who thinks in prompts, writes clearly, experiments fast, and cares about getting real results, do apply!
Lear Labs reposted this
Over the last month, we rolled out an AI-driven learning assistant with one of India’s largest test-prep platforms and the early results have been incredibly encouraging. In our first student feedback cycle (2500 responses), here’s what happened: ✨ 96%+ of students said the AI assistant felt knowledgeable, helpful, and engaging ✨ 55% are already using it to plan their daily preparation ✨ 90% would recommend it to their peers ✨ 72% gave it a perfect 5/5 ✨ And the favorite feature? AI-generated study plans — used by nearly half the students What excites me most is that students aren’t just liking or using the assistant, they’re actively demanding more AI-driven features, deeper personalization, and richer study plans. That level of trust and pull is huge. Our team at Lear Labs is now heads-down building the next version with stronger personalisation, richer academic intelligence, and new habit-building touchpoints. AI for learning is just getting started and I’ve never been more optimistic about where this is headed. 🚀 #AI #EdTech #LearningReinvented #PersonalizedLearning #AIProduct #FoundersJourney