Data Scientist

Data Scientist

Exp-3-8 Years

Mode-Remote

Np-Immediate-15 Days

Responsibilities

• Design, train, and evaluate ML models for tasks such as classification, anomaly detection, forecasting, and natural language understanding.
• Build and fine-tune deep learning models, including

  RNNs, GRUs, LSTMs

  , and

  Transformer architectures

  (e.g., BERT, T5, GPT).
• Develop and deploy

  Generative AI solutions

  , including

  RAG pipelines

  for use cases such as document search, Q&A, and summarization.
• Perform

  model optimization techniques

  such as

  quantization

  for improving latency and reducing memory/compute overhead in production.
• Optionally fine-tune LLMs using

  Supervised Fine-Tuning (SFT)

  and

  Parameter-Efficient Fine-Tuning (PEFT)

  methods like

  LoRA

  or

  QLoRA

  .
• Define and track relevant evaluation metrics; continuously monitor model drift and retrain models as needed.
• Collaborate with cross-functional teams (data engineering, backend, DevOps) to productionize models using CI/CD pipelines.
• Write clean, reproducible code and maintain proper versioning and documentation of experiments.

Requirements

Required Skills

• 4–5 years of hands-on experience in machine learning or data science roles.
• Proficient in Python and ML/DL libraries: scikit-learn, pandas, PyTorch, TensorFlow.
• Strong knowledge of traditional ML and deep learning, especially for sequence and NLP tasks.
• Experience with

  Transformer models

  and open-source LLMs (e.g., Hugging Face Transformers).
• Familiarity with

  Generative AI

  tools and

  RAG frameworks

  (e.g., LangChain, LlamaIndex).
• Experience in

  model quantization

  (e.g., dynamic/static quantization, INT8) and deployment on constrained environments.
• Knowledge of vector stores (e.g., FAISS, Pinecone, Azure AI Search), embeddings, and retrieval techniques.
• Proficiency in evaluating models using statistical and business metrics.
• Experience with

  model deployment

  ,

  monitoring

  , and performance tuning in production environments.
• Familiarity with Docker, MLflow, and CI/CD practices.