Senior Data Scientist

Senior Data Scientist

judgment and depth

Key Responsibilities

Model Strategy & Selection-

• Work with ML Engineers and Architects to

  evaluate and select LLMs and SLMs

  based on performance, cost, latency, privacy, and deployment constraints
• Define

  modeling approaches

  (prompting, RAG, fine-tuning, distillation) aligned to business use cases
• Establish benchmarking frameworks to compare open-source and proprietary models
  

Data Preparation & Training Readiness

• Partner with Data Engineering teams to

  design data pipelines

  for ingestion, cleaning, normalization, deduplication, and document chunking
• Define

  dataset standards, schemas, and quality metrics

  for pre-training and fine-tuning
• Ensure dataset lineage, versioning, and reproducibility across experiments
  

RLHF & Human-in-the-Loop Learning

• Design and lead

  RLHF / DPO pipelines

  , including preference collection and reward modeling
• Define annotation guidelines, scoring rubrics, and gold datasets
  
• Measure and improve data quality using inter-annotator agreement and bias metrics
  
• Continuously refine feedback loops to improve model alignment and reliability
  

Evaluation & Governance

• Define

  model evaluation frameworks

  covering relevance, hallucination, bias, robustness, and safety
• Support responsible AI practices and alignment with enterprise governance requirements
  
• Collaborate with product and business teams to translate outcomes into measurable impact
  

Technical Leadership

• Provide

  technical guidance and mentorship

  to data scientists and analysts
• Review experiments, data quality, and evaluation results
  
• Act as a

  bridge between data, engineering, and product teams

Required Skills & Experience

• Strong foundation in

  Data Science, Statistics, and Applied Machine Learning

• Hands-on experience with

  LLMs / SLMs

  , including fine-tuning and evaluation
• Expertise in

  Python, SQL, and large-scale text data processing

• Practical experience with

  RLHF, DPO, preference learning, or reward modeling

• Familiarity with

  PyTorch, Hugging Face, MLflow/W&B, and annotation tools

• Proven ability to operate in

  cross-functional, fast-moving environments