Senior Data Scientist

iSoftStone, Inc. is seeking a Senior Data Scientist to Join our Team!
*This is a Hybrid Role in the New York City Metro Area*
**Client Site Travel Required – Up to 25%**
***Candidates must have permanent authorization to work in the United States. Visa sponsorship is not available for this role.***
This role is a client-facing data scientist supporting enterprise retail accounts across merchandising, supply chain, customer, and pricing analytics. You'll work embedded with client teams — scoping the problem, building the model, and defending the methodology to business stakeholders who are not data people. This is a consulting role: billable, multi-account, and dependent on your ability to translate ambiguous business questions into tractable modeling problems.
Responsibilities:
Own end-to-end delivery on retail analytics engagements: discovery, data assessment, feature design, modeling, validation, deployment handoff, and results readout.
Build and productionize models across the retail value chain — demand forecasting and inventory optimization, customer segmentation and CLV, price/promo elasticity and markdown optimization, assortment and allocation.
Design data models and semantic layers on client data platforms (Snowflake, Databricks, Fabric, BigQuery); work with data engineering to define the tables the models actually need rather than accepting what exists.
Interrogate data quality and business logic before modeling — retail data is messy, and identifying the flaw in a returns table or a channel attribution rule is often worth more than a better algorithm.
Present findings to director- and VP-level client stakeholders; quantify business impact in margin, sell-through, GMROI, or working capital terms, not model metrics.
Support pre-sales: solution shaping, estimation, POC design, and technical credibility in client pitches.
Mentor junior analysts and contribute reusable accelerators to the retail practice.
Qualifications:
Five+ years applied data science experience, with meaningful time on retail, CPG, or e-commerce problems.
Strong data modeling fundamentals — dimensional modeling, star/snowflake schemas, slowly changing dimensions, grain definition. You should be able to look at a retail transaction feed and design the model, not just query it.
Advanced SQL and production-grade Python (pandas, scikit-learn, statsmodels); comfort with at least one of PyTorch/TensorFlow, Prophet/ARIMA-family forecasting, or causal inference frameworks.
Demonstrated experience with time series forecasting and/or econometric modeling (elasticity, uplift, incrementality).