AuraML
Category: Robotics / Embodied AI
Developer of a generative simulation platform (AuraSim) for training and validating robotics and physical AI systems using multimodal world simulation models built on NVIDIA AI infrastructure. AuraML was founded in 2022. The company is led by Ayush Sharma. Based in Bengaluru, Karnataka, India. Team size: 1-10. Total funding raised: $1.27M. Latest round: Pre-Seed. Key investors include Turbostart, Indian Angel Network (IAN), DeVC, GSF Accelerator.
- Founded
- 2022
- Headquarters
- Bengaluru, Karnataka, India
- Team size
- 1-10
- Total funding
- $1.27M
Value proposition
Generative simulation platform that reduces robotics testing from 6 months and $500K to 1 week at a fraction of the cost, by creating physics-accurate digital twins from multimodal inputs before any hardware deployment.
Products and solutions
AuraSim (Multimodal World Simulation Model) — generates physics-ready, high-fidelity simulation environments for robots from text descriptions, floorplans, videos, depth maps, and point clouds, Text-to-World generation, Sensor noise simulation (LiDAR, camera), Multi-robot deployment testing, Agentic automation pipeline for scenario variations
Unique value
India's first multimodal world simulation model for robotics built on NVIDIA Omniverse and Cosmos; enables simultaneous simulation of humanoid robots, AMRs, and robotic arms in shared virtual environments; agentic automation pipeline runs thousands of scenario variations without human intervention.
Target customer
Robotics companies developing autonomous mobile robots (AMRs), humanoid robots, industrial robotic arms; warehouse automation companies; logistics and manufacturing enterprises; physical AI researchers and developers.
Industries served
Warehouse Automation, Industrial Robotics, Manufacturing, Logistics, Autonomous Systems
Technology advantage
Built on NVIDIA Omniverse, NVIDIA Isaac Sim, and NVIDIA Cosmos world foundation models; first multimodal world simulation model from India leveraging NVIDIA AI infrastructure; proprietary synthetic data rendering engine; agentic automation for simulation pipelines; integration with ROS and existing robotics workflows.
How they differentiate
First multimodal world simulation model from India built on NVIDIA Omniverse/Cosmos; supports diverse input types (text, floorplans, videos, depth maps, point clouds); agentic automation pipeline for autonomous scenario generation; reduces testing time from 6 months to 1 week.
Main competitors
PsiBot, RoboDK, EIDON, Hillbot
Key partnerships
NVIDIA Inception, NVIDIA Omniverse, NVIDIA Isaac Sim, NVIDIA Cosmos, AWS, Google Cloud
Major milestones
Founded 2022, Raised $230K pre-seed from IAN (May 2023), Raised $1M pre-seed led by Turbostart (June 2025), Launched AuraSim multimodal world simulation model on NVIDIA AI infrastructure (Feb 2026), Launched Global Physical AI & Robotics Cohort with NVIDIA Inception (Feb 2026)
Growth metrics
Revenue: ₹40.6K (FY25); Employee count: 8; Web traffic: 1.11K monthly visits
Market positioning
Early-stage deeptech startup positioning as a critical infrastructure layer for the global robotics industry, with a focus on generative simulation for physical AI. Competes with established robotics simulation platforms by offering multimodal input flexibility and agentic automation.
Geographic focus
India (headquarters); expanding to US market
About Ayush Sharma
Ex-Rapyuta Robotics (Robotics Software Engineer); Ex-SAGA ROBOTICS/THORVALD (Autonomy Team); Ex-The Hi-Tech Robotic Systemz; Ex-Dimension NXG; Teaching Assistant at Northwestern University; Antler India Residency. Holds Master's in Robotics from Northwestern University (2015-2017); research/internships at IIIT Hyderabad.
Latest news about AuraML
More Robotics / Embodied AI companies
Official website: https://www.auraml.com
