Ai Data Science Associate Director - Blend360

Job Description
You will work as part of our global Data Science team to provide data driven AI solutions for our customers using state-of-the-artmethods and tools.
Work with practice leaders and clients to understand business problems, industry context, data sources, potential risks,and constraints
Work with practice leaders to getstakeholderfeedback, get alignment on approaches, deliverables, androadmaps
Create and maintainefficientdata pipelines, often within clients’architecture. Typically, data arefrom a wide variety of sources, internal and external, and manipulatedusing SQL, spark,andCloudbig data technologies
Assemble large, complex data setsfrom client and external sourcesthat meet functional business requirements.
Build analytics tools to provide actionable insights into customer acquisition, operational efficiency,and other key business performance metrics.
Perform data cleaning/hygiene, data QC, and integrate data from both client internal and external data sources on Advanced Data Science Platform. Be able tosummarize and describe data and data issues
Utilize deep learning principles and architectures, including CNNs, RNNs, and transformers, apply these techniques to natural language processing tasks.
Manipulate model parameters to achieve desired outcomes in text generation.
Crafteffective prompts that guide AI models to generate desired outputs. Understand how different prompt structures influence AI behavior.
Use RAG models, and combine a retrieval component with a generator to enhance the quality and relevance of the AI's output. Understand how to effectively integrate external knowledge sources into AI responses.
Train and fine-tune models on specific datasets to improve performance and ensure the relevance of the outputs to the task at hand.
Conduct statistical data analysis, including exploratory data analysis, data mining, and document key insights and findingstoward decision making
Document predictive models/machine learning results that can be incorporated into client-deliverable documentation
Assist client todeploymodels and algorithmswithin their own architecture