Spotlight

What if you could program a humanoid robot just by speaking to it?

Quick Pitch: Lightberry builds the conversational intelligence layer for humanoid and service robots. Its software lets robots hear, see, speak, and respond naturally without requiring developers or end users to write code.

The Problem

• Limited Intelligence: Most robots struggle with natural speech, perception, and contextual decision making.

• Complex Programming: Robot configuration requires specialized coding skills, slowing adoption beyond experts.

• Fragmented Systems: Manufacturers build siloed solutions that do not scale or interoperate across platforms.

Snapshot

• Industry: Robotics software and conversational AI

• Headquarters: San Francisco, CA

• Year Founded: 2025 (YC F25)

• Traction: Deployed at conferences, offices, and homes with early OEM partnerships

Founder Profiles

• Ali Attar, Co-Founder: Repeat YC founder (SigmaOS, YC S21)  with deep experience in human-computer interaction and consumer UX applied to robotics.

• Stephan Koenigstorfer, Co-Founder: CERN trained physicist and roboticist with a PhD. Deep expertise in real time control systems and embedded robotics.

Funding

• Current Round: Raising $3M (Seed)

• Lead Investors: Y Combinator 

• Other Backers: Pioneer Fund, multiple customers

• Total Raised: $500K (Pre-Seed)

Revenue Engine

• OEM partnerships: Pre installed social intelligence layer on robot hardware

• Platform licensing: Software licensed to robot manufacturers

• Open SDK: Developer-built skills expand use cases and adoption

What Users Love

• Voice based robot configuration without coding

• Vision, speech, and navigation integrated into one system

• Customizable personalities and discoverable skills

• Works across multiple robot manufacturers

Playing Field

• In-house OEM software: Narrow, manufacturer specific solutions

• Generic LLM integrations: Not designed for physical world interaction

• Early robotics platforms: Limited distribution and ecosystem depth

Why It Matters

Humanoid robots are transitioning from labs to real world environments. Hardware is advancing quickly, but adoption is constrained by software that cannot interact naturally or operate autonomously around people. Social intelligence is becoming the gating layer.

What Sets Them Apart

• OEM distribution: Embedded partnerships enable default deployment at shipment

• Purpose built architecture: Designed for perception, motion, and real world interaction

• Voice first configuration: Behavior defined conversationally, not through code

• Open developer platform: Enables third party skills and long term platform lock in

Analysis

Bulls Case 📈 

• OEM partnerships create early distribution leverage

• Repeat founders with complementary technical depth

• Clear platform positioning with ecosystem potential

• Natural language configuration lowers adoption barriers

• Developer extensibility scales value without linear cost

Bears Case 📉 

• Market timing for socially intelligent robots remains uncertain

• OEM partnerships may lack long term exclusivity

• Building a developer ecosystem requires sustained investment

• Performance depends on advances in AI models and robotics hardware

• Autonomous social interaction remains technically difficult in dynamic environments

Verdict

Lightberry targets the core constraint in robotics: human interaction. As hardware commoditizes, the control point shifts to software that enables robots to operate safely, intuitively, and autonomously around people. By focusing on social intelligence and OEM distribution, Lightberry sits between hardware and real world deployment. If humanoid robots scale commercially, interaction software is likely to define the winning platforms.

The Startup Pulse

• Moonshot AI — Raised 500 million dollars in a Series C led by IDG Capital, valuing the company at 4.3 billion dollars. Funding will expand AI infrastructure as it competes globally. Its Kimi K2 Thinking model is driving revenue and subscriber growth.

• Dazzle AI — The Palo Alto startup founded by Marissa Mayer raised 8 million dollars in seed funding at a 35 million dollar post money valuation. The round was led by Forerunner Ventures.

• Traini — Raised 7.5 million dollars to build a Cognitive Smart Collar that interprets dog behavior and emotional signals using AI.

Written by Ashher

Update your email preferences or unsubscribe here

© 2025 AngelsRound

228 Park Ave S, #29976, New York, New York 10003, United States