Jay Hack, an AI researcher specializing in natural language processing and computer vision, recognized several years ago that large language models (LLMs) like OpenAI’s GPT-4 or ChatGPT have the potential to enhance developer productivity by translating natural language requests into code.
After gaining experience as a machine learning engineer at Palantir and founding and selling Mira, an AI-powered shopping startup for cosmetics, Hack delved into experimenting with LLMs to automate pull requests— the process of merging new code changes with main project repositories. With the assistance of a small team, Hack gradually transformed these experiments into a platform called Codegen. Codegen aims to automate numerous mundane and repetitive software engineering tasks by leveraging LLMs.
In an email interview with TechCrunch, Hack explained, “Codegen automates the menial labor out of software engineering by empowering AI agents to ship code. The platform enables companies to move significantly quicker and eliminates costs from tech debt and maintenance, allowing companies to focus on product innovation.”
Codegen distinguishes itself from other code-generating AIs like GitHub Copilot, Amazon CodeWhisperer, and the Salesforce model by addressing “codebase-wide” issues such as large migrations and refactoring, rather than focusing solely on code autocompletion. Hack emphasized that Codegen employs a multi-agent system for complex code generation, orchestrating a swarm of agents to collaboratively decompose and solve large tasks, resulting in significantly better outputs.
The core product of Codegen is a cloud and on-premises tool that connects to codebases and project management boards like Jira and Linear. It automatically generates pull requests to address support tickets and can even set up some of the necessary code infrastructure and logging, according to Hack.
Codegen recently closed a $16 million seed round led by Thrive Capital, with participation from angel investors such as Quora CEO Adam D’Angelo and Instagram co-founder Mike Krieger. The funding brings Codegen’s total raised to $16.2 million and values the startup at $60 million post-money. Investors believe in Codegen’s potential to revolutionize software engineering tasks, freeing up developers to focus on innovation. While Codegen is currently incubating the platform with two “large-scale” enterprise partners and does not have paying customers yet, Hack anticipates growth in the coming year. The funds raised will be used to scale the workforce from six employees to 10 by the end of the year and support infrastructure expansion.
