Polar Announces Its Launch with a $5.7 Million Seed Round

Polar, the AI browser that actually does work for you, led by Kevin Jiang and Howard Zhong, has announced its launch with a $5.7 million seed round led by Madrona, with support from angels, including Thomas Dohmke (former CEO of GitHub), Erik Bernhardsson (Modal), Robert Yang (Fundamental), Nico (Fundamental), and many more.

Building an AI Browser for Autonomous Workflows.

Modern work is often divided between tasks that require human expertise, such as decision-making, creativity, and relationship building, and repetitive processes that consume significant time, including research, data entry, document updates, and administrative workflows.

Earlier generations of AI browsers focused primarily on assisting users with information retrieval and basic interactions across web pages. However, many workflows still required significant manual involvement, limiting the impact of these tools.

Drawing on experience building AI browser technology at Perplexity, the team behind the new platform identified an opportunity to create a more autonomous approach. Advances in AI models have enabled prototypes capable of completing longer, more complex workflows with greater reliability.

The vision is to create a browser where users can delegate extended tasks through simple instructions, enabling multiple AI agents to execute work in parallel and significantly expand individual productivity.

An AI Browser Built for Autonomous Web Tasks.

Polar is designed as an autonomous AI browser that can navigate websites, complete actions, and execute complex workflows on behalf of users. Operating through a user's existing browser environment, the platform can handle tasks ranging from short requests to extended multi-hour processes across different websites.

Over seven months, users have leveraged Polar to complete more than 4.5 million online actions across areas including research, recruiting, sales, operations, and marketing workflows.

The platform has gained adoption among individuals and teams through organic growth, becoming a recurring productivity tool for many users. Early customer feedback highlights significant time savings, with one company reporting more than 25 hours saved per week per employee after incorporating Polar into its workflow.

Benchmarking Long-Running AI Browser Agents.

Polar is positioning itself as an advanced browser agent designed to handle complex, long-running knowledge work tasks across the web. The company reports that the platform has achieved leading performance across several browser agent benchmarks, outperforming other tested AI systems.

According to the company, Polar achieved top scores on benchmarks including Odyssey and BU Bench V1, exceeding competing browser agents and large language models evaluated in the same categories. The company notes that existing benchmarks do not fully capture real-world knowledge work scenarios, where tasks often involve navigating complex websites and managing multi-step workflows.

The platform's performance is built around several technical capabilities, including tools for interacting with dynamic web environments such as spreadsheets, maps, documents, and social platforms. Polar can also coordinate multiple AI agents, assign tasks across agents, and combine different AI models to optimize performance across workflows.

A key component of the system is its memory architecture, which allows the platform to learn user preferences over time and improve its ability to complete recurring tasks. The company reports that Polar has successfully executed extended workflows lasting more than 15 hours.

A Vision for Automating Knowledge Work.

Polar's long-term ambition is to automate a significant share of knowledge work by creating AI systems that augment human capabilities rather than replace them. The company aims to help professionals spend more time on higher-value activities by delegating repetitive digital tasks to autonomous agents.

With much of modern knowledge work taking place through web-based tools and platforms, Polar is focused on building a general-purpose browser agent capable of operating across a wide range of websites and workflows.

The company envisions a future where AI agents become an essential part of everyday work, similar to how coding agents are reshaping software development. The goal is to make autonomous browser-based assistance a standard productivity layer for knowledge workers across industries.

Building the Infrastructure for AI-Powered Computer Use.

Recursive Intelligence is an applied AI company focused on developing end-to-end browser and computer-use agents designed to automate complex digital workflows.

The company was founded by a team with backgrounds spanning leading technology companies, financial institutions, and research organizations, including MIT, Y Combinator, Jane Street, Citadel, Apple, Perplexity, and Modal. The founding team brings experience across artificial intelligence, software engineering, and advanced problem-solving disciplines.

The company credits early users and supporters for helping shape the product's development, with early feedback playing a key role in refining its browser and computer-use technology.