Superhuman Coding Agents
Intro
The Problem
The Team
Proximal aims to give the world superhuman software engineering agents.
We are a data research lab for coding. Our goal is to build coding agents that can create their own high-quality training data, continuously and autonomously. We run experiments to discover what training data pushes capabilities forward, then build systems to generate them at a large scale.
01
The Problem
Doing this requires solving hard engineering and research problems:
What type of coding data will improve model capabilities?
How do we automate a meaningful portion of data creation using coding agents?
How do we build the infrastructure required to autonomously generate data at scale?
How do we create datasets spanning all of software engineering including external dependencies, infrastructure and more?
These are just a few of the many problems we need to solve to reach our goal. There's a lot of work ahead.
02
The Team
Our founding team has worked on RL infrastructure and coding agents at companies like Cursor and Prime Intellect and has previously exited companies.
We are backed by top investors and angels from OpenAI, Anthropic, Thinking Machines, xAI, Meta Superintelligence, Cursor and Cognition.
If this sounds exciting to you, consider joining us.
Intro
The Problem
The Team
Proximal aims to give the world superhuman software engineering agents.
We are a data research lab for coding. Our goal is to build coding agents that can create their own high-quality training data, continuously and autonomously. We run experiments to discover what training data pushes capabilities forward, then build systems to generate them at a large scale.
01
The Problem
Doing this requires solving hard engineering and research problems:
What type of coding data will improve model capabilities?
How do we automate a meaningful portion of data creation using coding agents?
How do we build the infrastructure required to autonomously generate data at scale?
How do we create datasets spanning all of software engineering including external dependencies, infrastructure and more?
These are just a few of the many problems we need to solve to reach our goal. There's a lot of work ahead.
02
The Team
Our founding team has worked on RL infrastructure and coding agents at companies like Cursor and Prime Intellect and has previously exited companies.
We are backed by top investors and angels from OpenAI, Anthropic, Thinking Machines, xAI, Meta Superintelligence, Cursor and Cognition.
If this sounds exciting to you, consider joining us.
