A new artificial intelligence system, Macroscope, has been unveiled by a team of seasoned entrepreneurs, including former Twitter head of product Kayvon Beykpour. This venture focuses on empowering developers and product managers by offering sophisticated AI-driven capabilities to analyze code, identify errors, and provide comprehensive summaries of codebase modifications. The system is designed to drastically cut down on time spent in administrative tasks and meetings, redirecting valuable engineering effort towards actual development and innovation.
Macroscope was founded in July 2023 by Kayvon Beykpour, who serves as CEO, alongside his long-time collaborator Joe Bernstein, known for their previous ventures Periscope and Terriblyclever. They are joined by Rob Bishop, co-founder of Magic Pony Technology, which was acquired by Twitter. The founders' collective experience across various companies, from startups to large corporations like Twitter, has informed their understanding of the challenges engineers face daily. They envision Macroscope as the essential tool they wished they had during their earlier endeavors, addressing the pain points of managing complex codebases and large engineering teams.
One of the primary goals of Macroscope is to alleviate the burden of tracking work across various platforms such as JIRA, Linear, and spreadsheets. Beykpour emphasized that engineers often spend too much time in meetings and not enough on actual coding. Macroscope seeks to rectify this by integrating seamlessly into existing workflows, starting with a GitHub app that grants access to the codebase. Further integrations with tools like Slack, Linear, and JIRA can then be implemented to enhance functionality.
The core technology behind Macroscope involves 'code walking,' a process that utilizes the Abstract Syntax Tree (AST) to build a deep understanding of a customer's codebase. This contextual knowledge is then synergized with large language models (LLMs) to deliver its diverse range of features. Engineers can leverage Macroscope to pinpoint bugs in their pull requests (PRs), generate concise summaries of their PRs, obtain overviews of codebase changes, and pose natural language questions related to code research.
For product leaders, Macroscope offers real-time insights into product updates, productivity metrics, and answers to queries regarding product functionality, code, or development activities. This empowers them to make informed decisions about engineering resource allocation and team priorities. Beykpour highlighted the system's ability to answer complex questions without requiring deep technical knowledge, making it invaluable for leaders and new team members alike, significantly reducing the need to interrupt senior engineers for explanations.
While Macroscope distinguishes itself with a broad set of features, it does operate in the competitive arena of code review tools, competing with solutions like CodeRabbit, Cursor Bugbot, Graphite Diamond, and Greptile. However, Macroscope claims superior performance, having conducted internal benchmarks that showed its product catching 5% more bugs than its closest competitor while generating 75% fewer comments, demonstrating its efficiency and accuracy in identifying issues.
The San Francisco-based startup, comprising a team of 20, has successfully secured $30 million in Series A funding, led by Michael Mignano at Lightspeed, bringing its total funding to $40 million. Additional investors include Adverb, Thrive Capital, and Google Ventures. Macroscope operates on a subscription model, costing $30 per active developer per month, with a minimum of five seats, and offers tailored enterprise pricing for larger organizations. It currently requires integration with GitHub Cloud.
Macroscope's innovative AI-powered platform is setting a new standard for code understanding and bug detection in the software development industry. By providing intelligent summaries, insightful analyses, and efficient bug identification, the system promises to significantly boost developer productivity and streamline the product development lifecycle.
