Software Engineer
San Francisco / Paris|$125,000 - $175,000
We are looking for a software engineer who ships fast and breaks conventions. You will work directly on the core product, building the infrastructure and interfaces that let developers run entire software factories from their phone.
This is not a role where you maintain legacy systems or attend status meetings. You will write code that matters, every single day. The codebase moves fast. If something is wrong, you fix it. If something is missing, you build it.
What you will do
- Build and own critical product surfaces across the full stack (TypeScript, React, Node.js)
- Design and implement systems for orchestrating cloud-based coding agents
- Ship features end-to-end, from database schema to pixel-perfect UI
- Work directly with founders to define what gets built next
- Operate with extreme autonomy and a bias toward action
You may be a good fit if
- You have built and shipped real products that people use
- You are mass-productive in TypeScript and can move across the stack without friction
- You have strong opinions about software architecture but hold them loosely
- You thrive in environments where the spec is a conversation, not a document
- You can debug a production issue at 2am and still push a feature by noon
- You care more about impact than process
Agent Research Engineer
San Francisco / Paris|$125,000 - $175,000
We are looking for someone obsessed with making AI agents actually work. You will design, build, and evaluate the agent systems at the core of Ascii, pushing the boundaries of what autonomous coding agents can do in real production environments.
This is a role where you think about agent architectures, prompt strategies, tool use, multi-agent orchestration, and failure modes. You will ship research into production, not write papers. Every improvement you make directly changes how thousands of developers build software.
What you will do
- Design and iterate on agent architectures for autonomous software development
- Build evaluation frameworks to measure and improve agent reliability and capability
- Develop prompt engineering strategies and tool-use patterns for coding agents
- Implement multi-agent orchestration systems that coordinate work across machines
- Research and integrate new LLM capabilities as they become available
You may be a good fit if
- You have hands-on experience building systems with LLMs beyond simple chat wrappers
- You understand the failure modes of AI agents and have opinions on how to fix them
- You can write production-quality code, not just notebooks and prototypes
- You follow the AI research landscape closely and can separate signal from noise
- You are comfortable with ambiguity and can design experiments to resolve it
- You care about making agents reliable, not just impressive in demos