Zeroset, an AI startup co-founded by Akshat Kannan and William Zhang, has raised $5.2 million in pre-seed funding. The company's first product, Nebula, aims to help AI agents understand and adapt to company workflows.
TL;DR
- Zeroset raises $5.2M to develop AI agents that learn from company workflows.
- Nebula tracks tasks, decisions, and activities across business systems to improve AI agent performance.
- The funding will support hiring, early deployments, and model training, with one run costing around $100,000.
What happened
Zeroset, co-founded by former Stanford student Akshat Kannan and University of Texas dropout William Zhang, emerged from stealth in October 2026 with $5.2 million in pre-seed funding. The round was co-led by Gradient Ventures and 2048 Ventures, with participation from Leblon Capital.
The startup's first product, Nebula, collects information from workplace software and tracks how tasks, decisions, and activities change over time. This helps AI agents understand the steps involved in a task, rather than simply searching through company documents for answers.
Zeroset plans to use the funding to hire researchers and engineers, support early deployments with businesses, and train its AI models. One training run costs about $100,000 over two weeks, according to Kannan.
Why it matters
Zeroset aims to address the challenge of automating multi-step tasks with AI agents, which often struggle with the unique processes and approval systems of each organization. By understanding these differences, AI agents can complete tasks with less human involvement.
The startup is entering a competitive field, with companies like Mem0 and Zep also developing technology to help AI systems retain and retrieve information. Zeroset's focus on workflow tracking sets it apart from other players in the market.
For developers and startups, Zeroset's technology could provide a valuable tool for building AI agents that can adapt to different company workflows. Investors may see potential in the startup's unique approach to improving AI agent performance.
Key facts
- Zeroset was co-founded by Akshat Kannan and William Zhang in 2026.
- The startup raised $5.2 million in pre-seed funding co-led by Gradient Ventures and 2048 Ventures.
- Nebula, Zeroset's first product, tracks tasks, decisions, and activities across business systems.
- The funding will support hiring, early deployments, and model training, with one run costing around $100,000.
- Zeroset had five employees when it announced the funding.
- Nebula is currently in a closed research preview, with access expected to expand gradually.
- The startup plans to combine license fees with usage-based charges for its products.
Context
Zeroset is part of a growing trend of startups focusing on AI agents and their ability to automate complex tasks. By understanding and adapting to company workflows, these agents can become more valuable tools for businesses.
The startup's focus on workflow tracking sets it apart from other players in the market, such as Mem0 and Zep, which are also developing technology to help AI systems retain and retrieve information.
As AI agents become more sophisticated, the need for tools that can help them understand and adapt to different company processes will only grow. Zeroset's technology could play a key role in this evolving landscape.
