Applied ML Research Scientist (PhD New Grad)

The Company
Flowstate is building the intelligence layer for social media content. Social media has become core marketing infrastructure ($40B category, growing +30% YoY), yet making high-performing creator content is still too slow, too manual, and too hard to scale. Flowstate turns brand context, creator context, and performance signals into creator-specific briefs, explainable feedback, and smarter creative decisions—so teams can create high-performing content faster without sacrificing authenticity, brand alignment, or creative quality. Every campaign strengthens the system, compounding into a powerful intelligence layer across creators, brands, and formats. For machine learning engineers, it’s a chance to build category-defining applied AI with real data, real feedback loops, and meaningful ownership from day one. 
The Role
In your first 4 months, you'll build a baseline multimodal LLM-based recommendation model with a mentor who has been a lead of a world-class recommendation system. From there, you'll push the system's capabilities across three fronts:

• Video Representation Learning: Design embedding models that extract and quantify the creative signals — pacing, structure, hook strength — that actually drive engagement.
• Closed-Loop Optimization: Build RL frameworks that use augmented and real-world data to continuously retrain and sharpen model outputs, turning every campaign into a learning signal.
• Agentic Tooling: Develop internal ML agents that accelerate your own research workflows — automated experimentation, eval pipelines, and model iteration at startup speed.

This system is a major component of Flowstate's core intelligence layer. What you build here compounds directly into the defensibility of our platform. You'll also serve as a ML subject matter expert across the team — partnering with AI engineers and product to translate research into production systems that empower creators and brands.
Minimum Qualifications

• Education: Current Ph.D. candidate or recent graduate in Computer Science, Electrical Engineering, or a related quantitative field (Expected graduation by end of 2026).
• Coding: Fluency in Python and deep learning frameworks such as PyTorch
• Research Track Record: One or more publications or presentations in top-tier ML conferences (e.g., ICML, CVPR, NeurIPS, ICLR).
• Location: Currently accepting candidates located in the USA only.

Preferred Qualifications

• Domain Expertise: Prior experience in building recommendation systems, multi-modal LLMs (VLMs), computer vision, or reinforcement learning.
• Research Versatility: While you have deep expertise in a specific area, you possess the "generalist" mindset required to solve problems across the broader ML stack.
• Agility: Prior experience in a startup or high-growth environment where research moves at a rapid execution pace.

Other Information

• Work Environment: Currently remote with a plan to transition to a hybrid model in NYC.
• Compensation: Base salary of $140,000 – $160,000 annually + annual performance bonus + sizable pre-IPO equity. 
• Sponsorship: H-1B sponsorship is available for qualified candidates, subject to company policy and legal eligibility.