Simular Inc.
Simular aims to free humans from computers and devices through AI agents, enabling people to live better, more human lives. The company builds autonomous computer agents that can navigate interfaces and complete workflows across desktops, browsers, and smartphones, freeing knowledge workers from repetitive computer work.
Founding Story
Simular was founded in 2023 by former DeepMind scientists Ang Li and Jiachen Yang who envisaged a shared vision: to harmonize human and AI intelligence, creating technology that empowers rather than replaces. The company emerged from the founders' extensive research background in reinforcement learning, autonomous agents, and computer vision, with the team researching AI agents since 2017. Their goal is to develop a neuro-symbolic continual reinforcement learning framework for building agents that work like humans, ultimately achieving AGI through autonomous computer use.
Discussions
No discussions yet
Be the first to start a discussion about Simular Inc.
Leadership
Founders
Ang Li
CEO and Co-Founder. PhD in Computer Science from University of Maryland College Park (thesis on spatial perception from visual and linguistic information). Previously worked at Google DeepMind (Waymo autonomous driving, Google Maps ETA prediction, resource efficiency), Baidu Apollo (autonomous driving as Principal Scientist), Facebook AI Research (Visual N-Grams), CMU Robotics, Apple, Google Streetview, and Comcast Labs DC. ACM/ICPC World Finalist. Over 40 publications in venues such as CVPR, ICCV, ECCV, NeurIPS, ICML, and ICLR.
Jiachen Yang (JC Yang)
Co-Founder and CTO. PhD in Machine Learning from Georgia Institute of Technology (dissertation on Cooperation in Multi-Agent Reinforcement Learning, 2021), supervised by Prof. Hongyuan Zha and Prof. Tuo Zhao. M.S. in Computer Science from Georgia Tech, B.S. in Electrical Engineering and Computer Sciences (EECS) from UC Berkeley. Previously worked at Lawrence Livermore National Laboratory (Staff Scientist), DeepMind, Electronic Arts, and Honda Research Institute leading agent creation and research. Specialist in multi-agent deep reinforcement learning. US Patent holder for cooperative multi-goal, multi-agent, multi-stage reinforcement learning.
Executive Team
Ang Li
CEO and Co-Founder
Ex-DeepMind, ACM/ICPC World Finalist, PhD from University of Maryland
Jiachen Yang (JC Yang)
CTO and Co-Founder
Ex-DeepMind, PhD from Georgia Tech, specialist in Multi-agent RL
Business Model
Revenue Model
Subscription-based SaaS model with tiered pricing (Plus, Pro, Enterprise) for desktop agents and cloud-based agents. API usage-based model for computer automation. Additional revenue from enterprise licenses with dedicated support and co-marketing partnerships.
Pricing Tiers
10,000 reloadable credits, 1 developer agent, community support, $10 referral bonuses
Unlimited credits, 1 free deployment agent, zero data retention, priority engineering support, $100 referral bonuses
Unlimited agents, managed scaling, dedicated account/engineering team, co-marketing partner options, SOC2, HIPAA, SSO, RBAC
Public computer use (requires waiting in a queue)
Dedicated computer in the cloud
Target Markets
- Insurance companies
- Healthcare organizations
- Recruiting and HR departments
- Travel agencies
- Knowledge workers
- Digital operations teams
- Insurance industry workflow automation
- Recruiting and HR processes
- Travel booking and coordination
- Healthcare forms and compliance
- Personal scripting and automation
- Business operations automation
- Workers in insurance sector
- Recruiting professionals
- Travel industry workers
- Microsoft
History & Milestones
Invited by Microsoft as early adopter to pilot Windows 365 for Agents program
Raised $21.5 million Series A led by Felicis Ventures; launched Simular 1.0, first desktop-native AI agent for consumers
Agent S3 achieved 72.6% success rate on OSWorld benchmark, surpassing human baseline (72.36%) for the first time
Agent S won Best Paper Award at the Agentic AI for Science Workshop at ICLR 2025
Achieved 90.1% success rate on WebVoyager (Browser Use) benchmark
