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
