Data Scientist
Noida, Uttar PradeshFull-TimeMid-levelAI / Data Science
Job Responsibilities:
- Data Analysis and Research: Analyzing a large dataset with queries and scripts, extracting valuable signals out of noise, and producing actionable insights into how we could complete and improve a complex ML and bidding system.
- Simulation and Modelling: Validating and quantifying the efficiency and performance gain from hypotheses through rigorous simulation and modelling.
- Experimentation and Causal Inference: Developing a robust experiment design and metric framework, and providing reliable and unbiased insights for product and business decision making.
Basic Qualifications:
- Master's degree in a quantitative discipline or equivalent.
- 2+ years minimum professional experience.
- Distinctive problem-solving skills, good at articulating product questions, pulling data from large datasets and using statistics to arrive at a recommendation.
- Excellent verbal and written communication skills, with the ability to present information and analysis results effectively.
- Ability to build positive relationships within ShyftLabs and with our stakeholders, and work effectively with cross-functional partners in a global company.
- Must have a deep understanding of ML algorithms, ranging from classical methods (e.g., regression, random forests, k-means clustering) to advanced techniques such as gradient boosting (XGBoost, LightGBM, CatBoost), neural networks, and transformer-based architectures (e.g., sentence transformers, BERT variants).
- End-to-End Deployment: Proven experience building, training, and deploying ML models from scratch into production environments, including model lifecycle management (versioning, monitoring, and retraining).
- Scalability & Performance: Hands-on experience operationalizing models at scale, optimizing for performance, reliability, and cost efficiency in real-world production systems.
- Programming: Experience with Python, R, or other scripting language, and database language (e.g., SQL) or data manipulation (e.g., Pandas).
