Senior Applied ML Engineer, Evals & Data
About the role
Senior Applied ML Engineer focused on evaluation, data quality, and measurement of Cardboard's AI agent. Own quality metrics, evaluation datasets, offline/online evaluations, and release gates to improve the agent that powers an in-browser video editor.
Responsibilities
- Define quality standards and build trusted evaluation datasets from real product usage.
- Build offline and online evaluations, including automated checks and human review.
- Analyze model and agent failure patterns, then improve quality through better data, evaluation methods, model selection, and, where useful, fine-tuning.
- Add regression checks and release gates while tracking quality, latency, and cost.
- Solve these hard problems.
Qualifications
- Experience shipping and operating an LLM or agent system used by real customers.
- Strong software engineering skills in TypeScript or Python, with the ability to work across both.
- Experience building evaluations, datasets, experiments, or AI quality systems.
- Strong product judgment and the ability to turn unclear quality problems into measurable improvements.
Skills
Benefits
- Competitive salary and founding-team equity.
- Unlimited tokens across every AI model. Use whatever you want, as much as you want.
- A healthy budget for AI tools and any peripherals you need to do your best work.
About Cardboard
Agentic video editor
Cardboard is an agentic video editor for growth/marketing teams and serious creators who need to ship videos consistently. Instead of agencies, long timelines, and infinite review cycles, teams give cardboard a footage + a goal (“3 variants”, “30s hook”, “testimonials”) to get a strong first cut in minutes on which they can iterate quickly. Multimodal LLMs can finally reason over footage, and WebGPU/WebCodecs make a real NLE possible in the browser. For the first time, an “AI director” inside a professional editor can be a reality. We have to build two hard things at once: a high-performance editor in the browser and an agentic editor that is reliable for production use cases. Most startups do one; incumbents can’t rebuild their stack without breaking everything. Saksham and Ishan met in school and have known each other since 15 years. Saksham was Co-founder/CTO of Iterate AI (backed by EF) and posts content on social media; Ishan has spent ~5yr building memory-heavy, performance-critical browser apps at HackerRank (S11) and deeply understands browsers. He has built Hotspoter (5M+ downloads) when he was 14. Together, we’re unusually suited to build the editor core others avoid. Video is becoming the default distribution channel, and the team that iterates fastest compounds attention. Cardboard becomes the default workspace for video production, the way Canva/Figma became the workspace for design - because it’s collaborative, fast, and does the tedious parts for you.
More open roles at Cardboard
- Founding Design Engineer1mo agoHybrid (Bengaluru, San Francisco) · Engineering
- Fullstack Engineer1mo agoBengaluru · Engineering
- Founding Designer1mo agoHybrid (Bengaluru) · Design
- Backend Engineer1mo agoHybrid (Bengaluru, India) · Engineering