About Clear Demand : Clear Demand is the leader in AI-driven price and promotion optimization for retailers . Our platform transforms pricing from a challenge to a competitive advantage, helping retailers make smarter, data-backed decisions across the entire pricing lifecycle. By integrating competitive intelligence, pricing rules, and demand modelling , we enable retailers to maximize profit, drive growth, and enhance customer loyalty — all while maintaining pricing compliance and brand integrity. With Clear Demand, retailers stay ahead of the market, automate complex pricing decisions, and unlock new opportunities for growth.
Key Responsibilities :
- People management - Lead a team of software engineers, DS, DE, MLE, in the design, development, and delivery of software solutions.
- Program management - Strong program leader that has run program management functions to efficiently deliver ML projects to production and manage its operations.
- Work with Business stakeholders & customers in the Retail Business domain to execute the product vision using the power of AI / ML.
- Scope out the business requirements by performing necessary data-driven statistical analysis.
- Set goals and, objectives using proper business metrics and constraints.
- Conduct exploratory analysis on large volumes of data, understand the statistical shape, and use the right visuals to drive & present the analysis.
- Analyse and extract relevant information from large amounts of data and derive useful insights on a big-data scale.
- Create labelling manuals and work with labellers to manage ground truth data and perform feature engineering as needed.
- Work with software engineering teams, data engineers and ML operations team (Data Labellers, Auditors) to deliver production systems with your deep learning models.
- Select the right model, train, validate, test, optimise neural net models and keep improving our image and text processing models.
- Architecturally optimize the deep learning models for efficient inference, reduce latency, improve throughput, reduce memory footprint without sacrificing model accuracy.
- Establish scalable, efficient, automated processes for large scale data analyses, model development, model validation and model implementation.
- Create and enhance model monitoring system that could measure data distribution shifts, alert when model performance degrades in production.
- Streamline ML operations by envisioning human in the loop kind of workflows, collect necessary labels / audit information from these workflows / processes, that can feed into improved training and algorithm development process.
- Maintain multiple versions of the model and ensure the controlled release of models.
- Manage and mentor junior data scientists, providing guidance on best practices in data science methodologies and project execution.
- Lead cross-functional teams in the delivery of data-driven projects, ensuring alignment with business goals and timelines.
- Collaborate with stakeholders to define project objectives, deliverables, and timelines.
Qualifications & Experience :
MS / PhD from reputed institution with a delivery focus.5+ years of experience in data science, with a proven track record of delivering impactful data-driven solutions.Delivered AI / ML products / features to production.Seen the complete cycle from Scoping & analysis, Data Ops, Modelling, MLOps, Post deployment analysis.Experts in Supervised and Semi-Supervised learning techniques. Hands-on in ML Frameworks - Pytorch or TensorFlow.Hands-on in Deep learning models. Developed and fine-tuned Transformer based models. ( Input output metric, Sampling technique)Deep understanding of Transformers, GNN models and its related math & internals.Exhibit high coding standards and create production quality code with maximum efficiency.Hands-on in Data analysis & Data engineering skills involving Sqls, PySpark etc.Exposure to ML & Data services on the cloud – AWS, Azure, GCP Understanding internals of computer hardware - CPU, GPU, TPU is a plus.Can leverage the power of hardware accel to optimize the model execution — PyTorch Glow, cuDNN, is a plus