You'll own the core AI systems that power Gamma, elevating quality, evaluating new frontier models, and pushing into new capabilities and modalities. This work operates at massive scale, with over 1 million AI-generated presentations and 6 million AI images created every day.

Your job is to productize existing foundation models, not train new ones. You'll focus on prompting, evaluating, and fine-tuning for maximum performance across Gamma's product surface. You'll also launch new modalities like voice and video, build the evaluation infrastructure to measure what 'good' looks like, and own uptime, latency, and cost across our AI stack.

You'll work closely with engineering and product to ship improvements that millions of users feel immediately.

What you'll do

• Own Gamma's LLM and image prompts, measuring and continuously improving quality at scale across text, layout, and visual generation
• Develop complex prompts for new features using AI JSX, balancing creativity with reliability
• Build evaluation frameworks for prompts and models, combining quantitative metrics with qualitative feedback to create better test sets
• Drive the AI roadmap based on quality gaps, constantly evaluating new frontier models and methods
• Curate datasets for fine-tuning open-source models and launch new modalities like voice and video
• Own analytics, tracking, uptime, latency, and cost across Gamma's AI infrastructure

What you'll bring

• Proven track record pushing foundation models to their limits, with hands-on experience building and evaluating prompts at scale
• Proficiency in TypeScript and Python, with a software engineering foundation that lets you move fluidly between prompt engineering and production development
• Strong data instincts: experience writing evals, designing metrics, and translating qualitative feedback into measurable, actionable improvements
• Self-sufficient in gathering and cleaning data to inform prompt improvements and model evaluations
• Experience with modern LLMs and image models such as Flux and Imagen (Nice to have)
• Familiarity with AI tooling like AI JSX for prompting and Braintrust for evaluations (Nice to have)

Compensation range:

The base salary for this full-time position, which spans multiple internal levels depending on qualifications, ranges between $180K - $300K plus benefits & equity.

XML job scraping automation by YubHub