• You want complete ownership over green fields LLM engineering projects from research to deployment
• You've previously built LLM applications from scratch with real users either as a founder, founding engineer, or a principal engineer
• You like to ship fast but you know when technical debt needs to be tackled
• You take pride in building elegant, scalable software
• You want to be surrounded by a team who thrives on shipping and solving hard technical problems
• You like to think about the good and bad futures for humanity in a world where AI agents automate large parts of the economy

Qualifications

• Proven experience building, deploying, and maintaining LLM-based systems that operate at scale
• Experience with LLM evals. You have felt the pain of trying to evaluate LLM-based systems
• Ability to translate customer feedback into specific LLM pipeline improvements
• Bachelor’s degree or higher in Computer Science, Artificial Intelligence, Software Engineering, Informatics, or a closely related technical field
• (Preferred) Experience with browser orchestration (fleet management, hot pooling etc.)

Tech Stack

• Core Technologies: Go, TypesScript, Python, Playwright
• AI Tools: OpenAI, Anthropic, Computer Use, HuggingFace, Observability + Evals frameworks
• Infrastructure: GCP, Docker, Kubernetes, Terraform

What You'll Build

• Robust browser automation systems powered by advanced computer-use AI models, enabling reliable, scalable agents that interact visually with browsers.
• Innovative graph-based workflow frameworks leveraging LLMs for deterministic browser automation, including automated agent-building pipelines that convert textual and video descriptions into ready-to-use workflows.
• Automated feedback loops translating human insights directly into prompt and workflow updates, supported by evaluation frameworks to measure and enhance model, prompt, and workflow improvements.
• Backtesting infrastructure to thoroughly validate system enhancements prior to deployment.
• Deterministic guardrails ensuring safe, predictable execution, especially for high-risk internal systems and portals.
• Rigorous evaluation and integration of cutting-edge open-source Vision-Language-Action GUI models.
• Direct customer deployments facilitating rapid iteration driven by real-world feedback and performance data.

About Us We’re David & Joe, both AI engineers.

David has built everything from robots, algorithms for a space agency to AI agents. He studied at top universities in the Czech Republic, Singapore, Israel and UK, crowdfunding his tuition fees. David loves nature, tennis, his family dog (Ferdinand Albrecht II), and Czech pilsner.

Joe has been researching and building LLM-based systems since GPT-2. He designed and built Netcraft’s Scam Intelligence system, an LLM-powered “victim” which engages with scammers to extract their bank details. His agent runs in the wild at scale and disrupts hundreds of millions of dollars of fraud annually. Joe is an avid mountaineer and adventurer.

We both feel privileged to be alive at this pivotal moment in history as we observe the invention of non-human general intelligence. We’re excited to help integrate these generally intelligent systems into the global economy.

Not because it will be easy, but because it will be hard.

Join us.