Seattle startup Caddi, a product of the AI2 Incubator, is emerging from stealth mode today, announcing a successful $5 million seed funding round. This funding will support its ambitious goal: to automate basic business operations using generative AI. Caddi’s core innovation lies in its “automation by demonstration” (AI Automation) technology. Imagine creating AI training videos simply by recording your screen and narrating the process. This is precisely what Caddi enables.
Users can create these recordings for various tasks, including customer onboarding, email management, document organization, data entry, and invoicing. The system then intelligently transforms these recordings into automated workflows, effectively creating a “virtual employee” capable of handling these routine tasks.Initially, Caddi is focusing its efforts on the legal and financial sectors, recognizing the significant potential for automation in these industries.
The startup’s name itself reflects its mission. Just as a golf caddie manages the behind-the-scenes tasks, allowing the golfer to focus on the game, Caddi aims to handle tedious business operations, freeing professionals to concentrate on their core expertise.
Caddi’s Founder and it’s aim
Caddi was founded in August by CEO Alejandro Castellano and CTO Aditya Sastry, both seasoned entrepreneurs. Castellano, originally from Peru, brings a strong financial background, having managed a portfolio of companies and investments before earning a master’s degree in engineering from Cornell University.
He then joined AI2 Incubator, eager to build software in a supportive environment. He notes that the incubator, though now independent, “were AI before AI was cool,” giving them deep expertise in building companies in this space.Sastry, a data science specialist, has a proven track record of leading engineering teams at multiple startups, most recently as the director of engineering at AgentSync, an insurance tech company.
Founding engineer Dallas Slaughter completes the initial team. Caddi plans to expand to 15 employees within the next year, reflecting its growth ambitions.The $5 million seed round was led by Ubiquity Ventures, with participation from Seattle-based Founders’ Co-op and AI2 Incubator, demonstrating strong investor confidence in Caddi’s vision.
Sunil Nagaraj, a general partner at Ubiquity Ventures, is joining Caddi’s board, further strengthening the startup’s leadership. Nagaraj’s previous investments include Auth0, a Seattle startup acquired by Okta for $6.5 billion, highlighting his experience in identifying and supporting successful tech companies.
Caddi tested its technology with pilot users in late 2024 and is currently onboarding new customers. The company anticipates an open launch later this year, at which point its automation tools will be available for customers to use independently. Caddi plans to offer its platform as a subscription service, providing a recurring revenue model. While the generative AI automation space is rapidly evolving and competitive, Caddi differentiates itself from existing solutions.
What Differentiate Caddi from Other AI Automation
Traditional no-code solutions, such as Zapier and Make, require users to understand technical concepts. Caddi removes this barrier, making automation accessible to a wider audience. Other competitors, like UiPath, Automation Anywhere, and OpenAI’s Operator, automate workflows by mimicking a human user’s actions. This approach can be unreliable and pose risks in regulated industries.
Caddi, in contrast, uses AI to analyze the intent behind a user’s actions in a screen recording. It then generates code that interacts directly with application APIs, building a more robust and reliable automated process. This approach significantly reduces the chances of errors and minimizes compliance risks, a critical advantage in sectors like legal and finance.
In a world where professionals face increasing pressure to improve efficiency and cut costs with AI, concerns about accuracy and reliability are paramount. Caddi directly addresses these concerns, offering a solution that is “way more reliable” and “way more cost effective” while mitigating “hallucination or regulatory risk.”
