We research and build AI systems that are safe, reliable, and beneficial โ grounded in scientific rigor and guided by a deep commitment to humanity's long-term wellbeing.
Our interdisciplinary teams tackle the most challenging problems in AI โ from foundational theory to real-world deployment.
Developing large-scale language and multimodal models with improved capabilities, efficiency, and alignment with human values.
Learn more โResearching alignment techniques, interpretability methods, and robust evaluation frameworks to ensure AI systems behave as intended.
Learn more โBuilding AI systems that understand and generate India's 22 official languages and hundreds of regional dialects with cultural nuance.
Learn more โApplying AI to accelerate scientific discovery โ from drug design and climate modeling to materials science and genomics.
Learn more โDesigning compute-efficient architectures and training methods to democratize access to powerful AI across resource-constrained settings.
Learn more โStudying how humans and AI systems can work together effectively, building tools that augment rather than replace human expertise.
Learn more โFounded in Bengaluru in 2023, Lemeno AI brings together world-class researchers, engineers, and ethicists united by a single purpose: to ensure that artificial intelligence develops in a way that is safe and beneficial for all of humanity.
We believe India is uniquely positioned to contribute to global AI โ with its rich tradition of mathematics, its linguistic diversity, and its vast pool of talent. We're building the research infrastructure for India to be a leader in responsible AI development.
We publish openly, peer review carefully, and build on reproducible foundations.
We optimize for humanity's long-term wellbeing, not short-term metrics.
We build AI that works for everyone โ across languages, cultures, and economic backgrounds.
Peer-reviewed papers advancing the state of AI research.
We present a novel approach to reward modeling that scales to complex tasks while remaining interpretable, with evaluations across 12 Indian languages.
A new fine-tuning paradigm that achieves state-of-the-art performance on 8 Indian language benchmarks using 10ร fewer parameters than existing approaches.
We introduce BharatLM, the first holistic evaluation suite covering 22 official Indian languages across 40+ tasks including reasoning, generation, and cultural understanding.
World-class researchers and builders passionate about beneficial AI.
Former research scientist at DeepMind. PhD from IIT Bombay. Pioneer in scalable alignment research.
Previously at Google Brain. Specializes in multilingual NLP and foundation models. PhD from IISc Bangalore.
Leading our alignment and interpretability research. Previously at Anthropic. MTech from IIT Delhi.
We're looking for exceptional researchers, engineers, and thinkers who want to work on some of the most important problems of our time. We offer competitive compensation, visa sponsorship, and a research-first culture.
Work on scalable oversight, interpretability, and alignment research. PhD preferred.
Apply Now โBuild and scale our training infrastructure for large language models. 4+ years experience.
Apply Now โ6-month research internship focused on Indian language understanding. Open to MTech/PhD students.
Apply Now โApply AI to biological and chemical research problems. Background in computational biology or chemistry preferred.
Apply Now โDon't see a role that fits? We'd still love to hear from you.
Get in TouchWhether you're a researcher, potential collaborator, or just curious about our work โ we'd love to connect.