JD for Data Scientist / ML Engineer
Location : Gandhinagar
Experience : 2 to 6 Years
About the Role
We're looking for a Data Scientist / ML Engineer to build demand and sales forecasting capability for our enterprise supply-chain and distribution planning products. You'll work on real production forecasting problems — predicting product-level demand across a distribution network, handling seasonality and promotional effects, and solving for products with little or no sales history — and you'll be responsible for getting those forecasts into the hands of planners through the applications our engineering team builds.
This is a hands-on, build-from-scratch role: there's no existing forecasting pipeline to inherit. You'll define the approach, build it, and iterate with direct client feedback.
What You'll Do
• Design and build demand/sales forecasting models (time-series methods such as ARIMA/SARIMA, exponential smoothing, Prophet, as well as ML-based approaches like LightGBM/XGBoost) for distribution and replenishment planning use cases.
• Handle real-world forecasting challenges: seasonality, promotional and scheme effects, and cold-start forecasting for newly launched products with no sales history.
• Build data pipelines to ingest and clean sales, stock, and order data from enterprise ERP systems (SAP experience is a strong plus) and structure it for modeling.
• Integrate forecast outputs into downstream planning applications via APIs or scheduled jobs, working closely with backend developers.
• Collaborate with business analysts and client stakeholders to translate supply-chain domain rules and constraints into model features.
• Track and report model accuracy (e.g., MAPE, bias), and iterate models based on real business outcomes, not just offline metrics.
• Build simple dashboards/reports to explain forecast performance to non-technical business stakeholders.
• Document your modeling approach and keep pipelines reproducible and maintainable by the rest of the team.
What We're Looking For
• 2–6 years of experience in data science or ML, including at least one real-world forecasting or time-series project (academic projects alone are not sufficient).
• Strong Python skills — pandas, numpy, scikit-learn — and familiarity with forecasting libraries such as Prophet, statsmodels, or equivalent.
• Solid grasp of time-series fundamentals: trend, seasonality, and how to approach cold-start/no-history forecasting problems.
• Working SQL skills and comfort querying relational databases directly.
• Experience building or consuming APIs to get model outputs into real applications — this role isn't notebook-only.
• Bachelor's or Master's degree in Computer Science, Statistics, Data Science, or a related field.
• Comfortable working directly with client stakeholders and explaining modeling decisions in plain business language.
• Comfortable in a small-team, multiple-hats startup environment.
Nice to Have
• Exposure to SAP data structures or other ERP integrations.
• Background in supply chain, FMCG, retail, or distribution/logistics domains.
• Familiarity with MLOps basics — model versioning, monitoring, and retraining pipelines.
What We Offer
• Ownership of a forecasting capability from the ground up, not a maintenance role on someone else's pipeline.
• Direct exposure to enterprise clients and real production impact, not just internal experimentation.
• A fast-growing company with room to grow into a lead data science role as the practice expands.
• Competitive compensation, based on experience.
