Data Scientist

About Fusemachines

Fusemachines is a 10+ year old AI company, dedicated to delivering state-of-the-art AI products and solutions to a diverse range of industries. Founded by Sameer Maskey, Ph.D., an Adjunct Associate Professor at Columbia University, our company is on a steadfast mission to democratize AI and harness the power of global AI talent from underserved communities. With a robust presence in four countries and a dedicated team of over 400 full-time employees, we are committed to fostering AI transformation journeys for businesses worldwide. At Fusemachines, we not only bridge the gap between AI advancement and its global impact but also strive to deliver the most advanced technology solutions to the world.
Key Responsibilities

• Design, build, and validate time-series forecasting models to predict key business metrics 

• Experiment with and evaluate a diverse range of modeling techniques, including traditional ML models (e.g., XGBoost), statistical methods (e.g., Prophet), and advanced deep learning approaches (e.g., DeepAR, N-BEATS).

• Collaborate closely with senior data scientists to analyze business requirements, perform feature engineering, and iterate on model improvements.

• Work with data engineers to understand and utilize complex datasets from Snowflake

• Contribute to the deployment of the model in a cloud environment, ensuring it is scalable and maintainable.

• Take ownership of the long-term model lifecycle, including performance monitoring, retraining, and incorporating new data inputs as they become available.

• Translate complex model outputs into actionable insights and communicate findings to stakeholders across different business units.

• Proven professional experience as a Data Scientist with a strong emphasis on time-series analysis and forecasting.

• Hands-on experience building and deploying forecasting models using libraries/frameworks such as:

  • XGBoost

  • Prophet

  • Deep Learning frameworks (e.g., PyTorch, TensorFlow)

• Proficiency in Python and its core data science ecosystem (e.g., pandas, NumPy, scikit-learn).

• Solid understanding of the underlying statistical and machine learning principles of various forecasting techniques.

• Ability to work collaboratively within a distributed team and effectively partner with client-side data scientists and engineers.

• Strong problem-solving skills with the ability to navigate complex data landscapes and translate business challenges into quantitative models.

• Direct experience with advanced neural network architectures for time-series forecasting, such as DeepAR or N-BEATS.

• Familiarity with querying and working with data from data warehouses, particularly Snowflake.

• Experience with MLOps practices, including model deployment, monitoring, and maintenance in a cloud environment.

• Domain knowledge in finance, supply chain, or sales operations planning is a significant plus