Data Scientist

Skills

Description

• Machine Learning : SciPy, Sckit-learn, Pandas
• Technique : Hyperparameter Tuning, Imbalanced Data, Ensemble Learning, Boosting
• Algo : SVG, XGBoost, KNN, Clustering, Views, MassTer, Stata etc
• Natural Language Processing : LLM familiarity (BERT), Neural Networks (LSTM)
• Big Data Experience : Cloud Services (GCP, Microsoft Azure, AWS)
• Expert knowledge on Time series Forecasting, Python Libraries
• 6+ years of experience in MMM, measurement, A/B testing, marketing strategy & analytics
• Proficient coding skills (Python/SQL) and database knowledge
• Experience with advertising technology platforms Ad servers, DSPs, DMPs, etc.
• Proficient in SQL and Python Libraries (Tensor Flow, Keras, PyTorch

Responsibilities

• • Conducting the appropriate analytics techniques on available data. This will focus on Market Mix Modelling but will also include a wider range of techniques such as portfolio management and digital analytics.
• • Deriving meaningful insights from the results of a project
• • Running and delivering practical and implementable media optimization scenarios supporting the decision-making process for the planning teams
• • Communicating results to internal teams, and establishing buy-in of recommendations
• • Working within a team of 10-13 people across different Team to ensure project deliverables are produced on time.
• • This is a hands-on role: Collecting data from relevant sources, checking for accuracy and formatting for analytics
• • Developing strong working relationship with the account planning teams and foster a culture of information sharing
• • Showcase Media Mix Modelling and its benefits to planning teams to create agency wide awareness
• • Learning, implementing optimization tools in cooperation with the wider data science team
• • Build and enhance media mix models to connect the impact of marketing drivers and business short-term and long-term outcomes
• • Develop optimization and simulation algorithms to guide the marketing investment and allocation recommendations to stakeholders
• • Scale current Python modeling algorithm through parallel processing or other efficiency enhancement
• • Design and execute A/B test and other experiments
• • Explain complex modeling approaches in simple terms and develop compelling narratives that connect modeling results with business problems
• • Partner with cross-functional team to streamline data science solutions