Willow’s $4.5 Million Bet on Voice-Controlled Workflows
Willow has raised $4.5 million as co-founder Allan Guo builds the voice-dictation company after leaving Stanford and entering Y Combinator’s startup ecosystem. The company has moved from testing healthcare uses to serving customers including Uber, with its software aimed at turning speech into text and controlling broader computer workflows.
Sunday, Aug 23, 2026
Tracking: Y Combinator
Geography: Silicon Valley, San Francisco Bay Area, United States
1. Willow Raises $4.5 Million After Allan Guo Left Stanford for Y Combinator
Willow co-founder and CEO Allan Guo says the AI voice-dictation startup has raised $4.5 million and landed customers including Uber.
Guo, now 20, left Stanford during his freshman year at 18 to build the company with Lawrence Liu after entering Y Combinator’s startup ecosystem.
Willow began in healthcare, where the founders spent about a year testing ideas before observing professionals use voice AI for documentation.
Its software turns speech into polished text for emails, messages, documents and AI prompts, while the company describes a broader goal: voice control across computer applications and workflows.
In 2025, Guo announced a $4.2 million round led by BoxGroup; Goodwater Capital, Burst Capital and Liquid 2 Ventures participated.
Key facts:
Allan Guo left Stanford during his freshman year at 18.
Willow says it has raised $4.5 million and counts Uber among its customers.
BoxGroup led Willow’s $4.2 million 2025 funding round.
Goodwater Capital, Burst Capital and Liquid 2 Ventures participated.
Willow was co-founded by Guo and Lawrence Liu.
Why it matters: Willow shows how an unusually young founder used Y Combinator while iterating from healthcare software toward a broader voice-computing product.
The company’s reported financing and Uber customer offer concrete evidence of commercial progress, although the article provides no revenue or customer-count details.
The next question is whether Willow can expand beyond dictation into reliable voice control across workplace applications and AI tools.
That ambition would place it in a broader product category than transcription, but the article does not establish how widely those capabilities are used today.