Lead AI Engineer (Multimodality)

TL;DR

Lead AI Engineer (VLM/LLM): Building and owning the multimodality capabilities for frontier coding models with an accent on image input and production-grade VLM training. Focus on transitioning from adapter-based approaches to native multimodality and scaling training on thousands of GPUs.

Location: Remote (EMEA/East Coast)

Company

Poolside is an AI research and engineering company focused on reaching AGI by accelerating software development through agentic systems and open-weight frontier models.

What you will do

• Own the multimodality direction and capability adoption for Poolside's frontier models.
• Ship the first image-input capabilities, enabling models to read designs, specs, and diagrams.
• Chart and drive the technical path from adapter-based approaches to native multimodality.
• Collaborate on the creation of custom evaluations and datasets for multimodal SWE capabilities.
• Run end-to-end experiments, including hypothesis formulation, large-scale training, and analysis.
• Partner with evals, architecture, data, and post-training teams to integrate multimodality.

Requirements

• Experience in end-to-end training of production-grade VLMs.
• Strong fundamentals in LLM training, transformers, and distributed training at scale.
• Strong Python programming skills.
• High ownership and ability to thrive in greenfield ambiguity.
• Must be based in EMEA or US East Coast.

Nice to have

• Leadership and 0-1 experience with substantial breadth and hands-on contributions.
• Distinguished research background in VLMs or multimodal models.
• Deep understanding of native multimodality challenges.
• Experience building agentic computer use systems.

Culture & Benefits

• Fully remote work with flexible hours.
• 37 days per year of vacation and holidays.
• Health insurance allowance for employees and their dependents.
• 16 weeks of flexible, full-pay parental leave.
• Allowances for well-being, home office, and continuous learning.
• Frequent team get-togethers, including monthly 3-day sessions in Paris.

Hiring process

• Introductory call with the Co-Heads of Applied Research.
• Technical interviews with Founding Engineers and Engineering team members.
• Team fit discussion with the People team.
• Final interview with the Co-CEO and Co-Founder.