Skip to main content

Parallax Worlds

Category: Robotics / Embodied AI

A high-fidelity robotics simulation and reliability testing platform that utilizes AI to bridge the 'sim-to-real' gap by converting simple video captures into interactive 3D digital twins. Parallax Worlds was founded in 2024. The company is led by Tanmay Agarwal. Based in Santa Clara, USA. Team size: 1-10. Total funding raised: $4.9M. Latest round: Seed, $4.0M, 2025-11, led by Pear VC. Key investors include ["Pear VC","Kakao Ventures","GS Futures","Lightscape Partners","Unusual Ventures","Conviction","Boost VC"].

Founded
2024
Headquarters
Santa Clara, USA
Team size
1-10
Total funding
$4.9M

Value proposition

Enables companies to measure robot reliability upfront and detect failures in a virtual environment before physical deployment, significantly reducing on-site iteration time and deployment costs.

Products and solutions

["Video-to-3D Reconstruction Pipeline (iPhone-to-Digital Twin)","Modern Robotics Reliability Testing Platform","AI-Generated Stress-Testing Scenarios","Sim-to-Real Integration Suite"]

Unique value

Developed a proprietary pipeline that generates fully navigable, hyper-realistic digital twins from simple video walkthroughs (e.g., iPhone footage), eliminating the need for complex manual 3D modeling of environments.

Target customer

Robotics manufacturers, R&D teams, warehouse operators, and industrial automation startups (e.g., Orangewood Labs, BotBuilt, Rainier Labs).

Industries served

["Industrial Manufacturing","Logistics and Warehousing","Construction Robotics","Autonomous Systems","Deep Tech / AI Research"]

Technology advantage

Vertically integrated platform that supports real-world robot software execution within simulations; compatible with industry-standard engines like NVIDIA Omniverse, Unreal, and Unity while adding robotics-specific physics and failure-mode prediction.

How they differentiate

Parallax Worlds differentiates through its proprietary video-to-3D AI pipeline, which allows users to create hyper-realistic digital twins of factories and warehouses using only an iPhone. Unlike traditional simulation platforms that require manual 3D modeling, Parallax focuses on high-fidelity reliability testing and 'failure mode' detection to bridge the sim-to-real gap.

Main competitors

["NVIDIA (Isaac Sim)","Applied Intuition","Parallel Domain","Unity Robotics"]

Key partnerships

["NVIDIA (Omniverse integration)","Kakao Ventures (Strategic partner for Asian market entry)","Pear VC (Lead investor and Silicon Valley network partner)","Lightscape Partners"]

Notable customers

["Orangewood Labs","BotBuilt","Rainier Labs"]

Major milestones

["Incorporated as Dalus AI Inc. in early 2024","Successfully developed proprietary 'iPhone-to-Digital Twin' reconstruction technology","Announced $4M Seed round led by Pear VC in late 2025/early 2026","Strategic partnership for Asian market entry established with Kakao Ventures"]

Growth metrics

The company has transitioned from stealth to seed-stage growth, securing $4.9M in cumulative capital and onboarding several active robotics manufacturing customers within its first year of operation.

Market positioning

Seed-stage deep tech challenger in the robotics simulation and reliability testing market.

Geographic focus

North America (Silicon Valley headquarters) and Asia-Pacific (specifically South Korea via strategic investors like Kakao Ventures and GS Futures).

Patents and IP

Proprietary AI-driven video-to-3D reconstruction and failure analysis algorithms (No specific public patent numbers disclosed as of January 2025).

About Tanmay Agarwal

Tanmay Agarwal is a robotics and machine learning expert with a research background from Stanford University's Vision & Learning Lab (SVL). Under the mentorship of AI pioneers like Fei-Fei Li and Jiajun Wu, he focused on 3D scene graphs and autonomous systems. His work specifically addresses the 'sim-to-real' bottleneck in robotics, leveraging high-fidelity virtual environments to ensure real-world reliability.

Official website: