Are you someone who enjoys working on challenging and complex business problems? Do you have years of experience building large-scale, time-series forecasting models? Are you hands-on with machine learning and statistical methods implementation? Do you have a passion to add value to business processes and make impact on business user’s strategic decisions and ultimately impact millions of consumers around the world?
If your answer is yes to these questions, then we have an exciting opportunity within our fast-growing data science team in the US Data and Analytics organization at Kraft Heinz in the heart of the great city of Chicago. You will be responsible for developing time-series forecasting methodologies for our customer teams in the US and scaling it to different channels (big box, supermarkets, convenience stores, wholesalers). You will collaborate with different functions across our business (sales, finance, IT) to implement your solutions in an enterprise application. To succeed in this role, you should feel comfortable implementing machine learning (ML) solutions in Python and have experience with deploying ML models on Azure. In addition to your implementation skills, you should have demonstrated a proven track record of delivering solutions that had made an impact to the business.
The data scientist for this role would not only develop ML algorithms but is also expected to be the owner of model forecasting accuracy across all the item/customer combinations in the Kraft Heinz portfolio. This person will deep dive into forecast measurements, identify opportunities to improve data quality in drivers behind prediction, integrate these findings into well-engineered feature sets and test/validate predictions on an ongoing basis. To deliver the necessary business impact, this role will:
communicate findings cross-functionally integrate feedback from customer sales teams into his/her work drive adoption of driver-based forecasting approach in the company.
Develop state-of-the art machine learning algorithms in Python to improve existing forecast accuracies at scale using Azure tech stack. Enhance the current data science model to drive adoption by sales teams.
Collaborate with customer sales, finance and IT teams in driving adoption of model forecasts in our internal promotion planning application. Be a champion of the model in the organization, interpret model results to non-technical audiences and integrate actual users’ feedback in day-to-day model development, and facilitating change-management with key stakeholders
Master’s degree in quantitative field (computer science, industrial engineering, statistics, mathematics)
3 years of experience in time series forecasting using traditional and modern methods – ARIMA, SARIMAX, FBProphet , XGBoost, RNN-LSTM
3 years of experience in econometrics and advanced regression methods – generalized linear models, segmented regression, Logit, Probit, Lasso regression
Proficiency in Python
Experience with developing ML models on Cloud (Azure preferred)
Strong SQL database skills, familiarity with big data technologies (Apache Spark)
Minimum 2 years of experience delivering ML-based Cloud solutions on Microsoft Azure
2-3 years of industry experience as a technical leader in data science/ML/AI
5-6 years of programming experience (Python, R, Java, C++)
Maintain a Github repository, actively participate in data science competitions, mentoring
Consumer Product Goods (CPG)/retail industry experience
Experience with Nielsen/IRI data
Experience with predictive modeling in the promo space