ML Engineer

Responsibilities:

• You'll architect intelligent systems that sense demand shifts in real time and recommend actions before a human even asks.
• Agentic AI pipelines that go beyond prediction systems that reason, decide, and adapt autonomously.
• Advanced demand forecasting and sensing models using attribute-based and cohort-based approaches.
• Signal integration frameworks that fuse external data and real-time market inputs.
• Feature engineering at scale: lag structures, rolling aggregations, and derived features.
• Model performance frameworks tied directly to measurable business outcomes.
• Translating complex model insights into clear, actionable language for clients and leadership.

Key Performance Indicators:

• Model accuracy, reliability, and real-world forecast performance.
• Speed and quality of model deployment to production pipelines.
• Business impact of AI-driven recommendations measured in inventory efficiency and revenue outcomes.
• Cross-functional collaboration effectiveness with product, engineering, and client teams.

Requirements:

• You bring a strong foundation in machine learning and a genuine curiosity about where AI is heading.
• 2-6 years of hands-on experience in ML, AI, or Data Science with a proven model-building track record.
• Python fluency: pandas, scikit-learn, XGBoost, and LightGBM are second nature to you.
• Deep understanding of time series forecasting and experience with large-scale transactional datasets.
• Ability to connect model outputs to real business impact and articulate the 'so what' clearly.
• Genuine curiosity about the power of AI and openness to emerging tools and paradigms, including agentic and generative AI.

Bonus Points:

• Experience in retail, supply chain, or inventory planning domains.
• Exposure to hierarchical forecasting, new product forecasting, or ensemble methods.
• Hands-on work with LLM-based agents or orchestration frameworks.

Skills

AI, ML, artificial intelligence, data science, forecasting, lightgbm, machine learning, pandas, python, scikit-learn, time series, xgboost