Applied Scientist (Behavior & Modeling)

Imagine websites thatrebuild themselves on autopilotto sell more, convert better, and grow faster. AtShuttlebase, we’re making that happen merginggenerative AI, large-scale data pipelines, and experimentationinto a self-evolving web. We’re awell-funded, very early-stage startupwith a massive vision. Joining now means you’re not just an employee you’re helpingset the technical and cultural DNAof the company. If you’ve ever wanted the autonomy of a founder with the backing of top-tier investors, this is it. As Applied Scientist, Behavior Modeling, the company expects you to carry out the following tasks: Build the first version of Shuttlebase’s behavior model, which turns raw session events into embeddings of visitor behavior. You choose the approach. Pre-training a transformer on event sequences is a likely starting point. Work with short, anonymous sessions, where most visitors never come back. Use the model to detect buying intent during a session and group visitors into audiences, so the AI can show each group the right version of the page. Get a first version live on real stores quickly, then improve it step by step. Read new research, test promising methods on Shuttlebase’s data and decide what to adopt. Work daily with the engineers who build on top of the model, and with the client team that explains results to stores. You’re the person they’re looking for if...You have a PhD or equivalent research experience in a quantitative field (physics, math, computational biology, neuroscience or similar). An exceptional master’s is also considered (Required) You have trained models on messy, real-world sequence or time-series data and can show whether they worked (Required) You have a proven track record shipping ML models to production in industry (Required) You have working knowledge of sequence learning (transformers, self-supervised pre-training, embeddings), enough to choose between approaches and explain why (Required) You have solid Python and can pull your own data with SQL (Required) You can explain technical work to non-specialists, through talks or writing You stay on top of current research and know how to apply it to real products You are available to work onsite 4 days per week in Midtown Manhattan, NYC (Required) They’ll be even more excited about you if you also...Have experience with recommendation systems, user behavior data, large event datasets or eCommerce