Senior Data Scientist

meredith; United States; Hybrid; USD 150,000 - 230,000; Full time

Job summary

Meredith is seeking a Senior Data Scientist for personalization to own the science behind the recommendation engine powering each user’s personalized product feed. You will design, build, and evaluate models using signals from user saves and live catalogs, aiming to scale from hundreds of thousands to tens of millions of products.

This hands-on role covers end-to-end development from data to production, with an emphasis on cold-start challenges and rigorous experimentation to improve engagement

Qualifications

• Own the design and development of the core recommendation models that turn user‑saved data into a personalized feed.

Responsibilities

• Own the design and development of core recommendation models and multi-signal approaches.
• Define and run rigorous offline and online experiments (A/B and multivariate) and translate results into product decisions.
• Turn implicit behavior and catalog attributes into meaningful features using SQL, NLP, and computer‑vision techniques.
• Take models from prototype to production within the existing MLOps stack and monitor production performance.

Skills

Recommendation & Personalization Mod

Experimentation & Measurement

Data, Signals & Features

Full-Cycle Productionization

Job description

Job Title Senior Data Scientist Job Description About The Position

As a Senior Data Scientist for personalization, you will own the science behind the recommendation engine that powers each user's personalized product feed. Starting from our user-saved product signals and a live catalog ingested from thousands of retailer feeds, you will design, build, evaluate, and continuously improve the models that learn each user's taste across brand, category, color, price point, and fit.

This is a hands‑on, full‑cycle role. You will take a recommendation problem from raw data all the way to a production model running on our existing MLOps stack - you own the model layer, not the infrastructure. Our platform team already operates the feature store, serving, and autoscaling; your job is to decide what to model, prove it works through rigorous offline and online experimentation, ship it, and iterate as behavioral signals accumulate.

A defining challenge of this role is cold-start. We are launching with a small behavioral dataset and a catalog scaling from hundreds of thousands of products toward tens of millions. You will need strong commerce and product‑data intuition to produce high‑quality recommendations before rich click data exists - and the experimental discipline to keep improving them as it arrives. This is a foundational hire that will shape how millions of users discover products they love.

Remote or Hybrid 3x a week NYC In‑office Expectations: This position offers remote work flexibility; however, if you reside within a commutable distance our offices in New York, the expectation is to work from the office three days per week.

About The Team

Our next‑generation product discovery platform connects shoppers with the things they love across thousands of retail partners. Users save, organize, and share products they're excited about - and our platform turns those signals into a deeply personalized shopping experience. We ingest live product feeds from thousands of retailers and use a rich understanding of each user's taste to surface the right product at the right moment.

We're building the recommendation engine at the heart of this shopping experience - a system that understands not just what people save, but why they save it. This is a foundational hire that will shape how millions of users discover products they love.

About The Positions Contributions

• 35% Recommendation & Personalization Modeling

  Own the design and development of the core recommendation models that turn user‑saved product data into a personalized feed. Develop multi‑signal models spanning brand affinity, category, color/visual attributes, fit and sizing, price sensitivity, and trend. Select and justify approaches across collaborative filtering, matrix factorization, content‑based, and hybrid/neural methods (e.g., two‑tower and other embedding models), and know when each applies. Build product and user embeddings that capture semantic similarity across the catalog and power candidate generation and retrieval. Design cold‑start strategies that produce high‑quality recommendations for new users and newly ingested products with little or no behavioral history.

• 25% Experimentation & Measurement

  Define what "good" personalization means and how it is measured. Establish rigorous offline evaluation (ranking and relevance metrics, sound holdout design) and connect it to online outcomes. Design, run, and read out A/B and multivariate experiments, and translate results into clear product and business decisions. Bring statistical discipline - sound experiment design, awareness of bias and confounding, and honest interpretation - so the team can trust which changes actually move engagement.

• 20% Data, Signals & Feature Understanding

  Develop deep intuition for Picksy's product catalog and user signals. Turn implicit behavior (saves, clicks, dwell, shares) and catalog attributes into meaningful model features, writing SQL against BigQuery to pull, join, and shape raw data into training/evaluation datasets. Apply NLP and computer‑vision techniques - including modern embedding and LLM‑based approaches - to extract structured attributes (category, color, material, fit) from unstructured product descriptions and imagery, and to enrich sparse catalog data. Partner with data engineering on data quality, freshness, and coverage as the catalog scales from hundreds of thousands toward tens of millions of products.

• 20% Full‑Cycle Ownership & Productionization

  Take models from prototype to production yourself. Write clean, production‑quality code and deploy into the existing MLOps pipeline (feature store, training, serving, monitoring) rather than building infrastructure from scratch. Own model performance in production: instrument it, watch for drift and degradation, and iterate as behavioral signals accumulate. Document models, features, and decisions clearly, and collabora