Wordware raises $30M to re-imagine AI development as ordinary writing
Wordware, a San Francisco startup looking to make programming artificial intelligence agents as simple as writing ordinary English sentences, today announced it raised $30 million in seed funding led by Spark Capital.
The investment also attracted participation from Felicis, Y-Combinator, Day One Ventures, and notable angel investors including co-founder and Chief Executive of Runway Siqi Chen, co-founder of Webflow Vlad Magdalin and co-founder and CEO of Ironclad Jason Boehmig.
Today, the development process for AI software and agents is led by software engineers and experts who can fine-tune prompts to get models to perform exactly how they want. Wordware argues that the industry is stuck between a rock and a hard place when it comes to tools, either AI logic is stuck in the codebase or it's built using no-code solutions that can't help cover every angle.
This leaves a gap where experts who have the ideas in their heads and understand the industries where AI will be used but cannot express concepts clearly for AI models and agents to act on. Their knowledge more often than not gets lost in translation when it's passed on.
"AI development needs a ground-up reimagining," said Filip Kozera, Chief Executive of Wordware. "This vision drove us to spend a year rebuilding the development environment for AI agents, solving the fundamental question: 'If prompting is the new programming, what should the tools look like?'"
AI agents built by Wordware's platform use English as a programming language. Users do not need to understand any complex prompting and if they can express their ideas clearly, they can build complex and sophisticated AI solutions.
The company said that it intends to set a balance between advanced no-code editors and natural language prompt systems. It allows everyday users to connect ordinary English language descriptions of goals for AI agents to different AI models and tools such as speech recognition, speech synthesis, image generation or data analysis.
"Until now assembling an AI meant building an entire engineering team, months of coding and days needed for each iteration. But that changes today," said Kozera. "We've created something entirely new. Now English is the programming language; accessible for everyone, yet powerful enough to satisfy your CTO."