We are looking for a skilled and hands-on Data Scientist with 3–8 years of experience in developing and deploying machine learning models—ranging from traditional ML algorithms to advanced deep learning and Generative AI systems. The ideal candidate brings a strong foundation in classification, anomaly detection, and time-series modeling, along with hands-on experience in deploying and optimizing Transformer-based models. Familiarity with quantization, fine-tuning, and RAG (Retrieval-Augmented Generation) is highly desirable.

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.