AI Applied Scientist
Drive AI research and build generative AI models using techniques like SFT, RL, and prompt engineering to enhance Figma products. Requires 4+ years in Generative AI, 6+ years in ML/NLP/CV, and strong Python skills.
About the job
What you’ll do at Figma:
- Drive fundamental and applied research in AI, exploring boundaries of current technology to build best-in-class models for Figma’s domains.
- Combine industry best practices and first-principles to build cutting-edge Generative AI models using techniques like Supervised Finetuning (SFT), Reinforcement Learning (RL), prompt improvements, and synthetic data generation.
- Work with product and infrastructure engineers to improve Figma’s products through AI-powered features.
- Collaborate with product managers and engineers to transform user feedback into AI system requirements.
- Build evaluation systems to measure and improve quality of AI features in Figma products.
We'd love to hear from you if you have:
- Extensive experience building generative AI features through prompt engineering and fine-tuning models in production.
- Experience with deep learning and generative AI frameworks like PyTorch, JAX, HuggingFace.
- Experience training LLMs with Reinforcement Learning techniques such as preference-based RL (DPO, PPO) and/or RL with verifiable rewards (RLVR) like GRPO/DAPO.
- 4+ years in Generative AI, and 6+ years in machine learning, natural language processing/understanding, or computer vision.
- Strong software engineering skills with 5+ years in programming languages (Python, C++, Java, or R).
- Experience communicating and working across functions to drive solutions.
Nice-to-haves:
- Proven track record of planning multi-year roadmaps.
- Experience mentoring/influencing senior engineers.
- Expertise in large-scale and distributed AI training.
Skills
PyTorch, JAX, Huggingface, Generative AI, Reinforcement Learning, LLMs, Prompt Engineering, Supervised Finetuning, Dpo, Ppo
Similar jobs
Data Science jobs
Confido
New York, NY
Build and ship forecasting, promotion-effectiveness, and data-quality models using retail and CPG data. The role requires 3–5 years of data science experience, strong statistical foundations, and fluency in Python and SQL.
Notion
San Francisco, CA
The Data Scientist will use experimentation, statistical inference, and data analysis to guide Growth strategy, define metrics, and evaluate product opportunities. The role partners with product, engineering, and marketing leaders and requires strong SQL plus Python or R skills.
Ramp
New York, NYOwn Ramp’s AI-focused economic research and build a customer-facing economics function using large-scale business spend data. The role requires 5+ years as a data scientist or economist, strong SQL and BI expertise, rigorous writing, data visualization, and executive communication.
PrizePicks
United States
Leads the data science function responsible for sports and esports market origination, projection accuracy, and coverage strategy. Requires 5+ years in sports betting, fantasy, or gaming, strong modeling skills in SQL and Python or R, and experience scaling high-performing teams.
Haus
San Francisco, CAOwn customer-facing causal marketing measurement engagements, from model configuration and QA through interpretation, recommendations, escalations, and enablement. The role requires direct MMM experience, marketing experimentation knowledge, and strong customer communication skills.