Description : Lead Analyst, AI Engineer (Data Science)

Experience : 8 to 12 Years

Location : Chennai

Education : Bachelors (7+ yrs exp), Masters (5+ yrs exp), or Doctorate (2+ yrs exp) in a quantitative field.

Role Summary

We are seeking a high-caliber Lead Analyst and AI Engineer to join our data science leadership team in Chennai.

In this role, you will act as a "Technical Architect of Intelligence," responsible for designing, prototyping, and deploying enterprise-grade AI systems.

You will bridge the gap between traditional Machine Learning and cutting-edge Generative AI, leveraging frameworks like PyTorch and Hugging Face to build scalable solutions.

The ideal candidate is a hands-on expert who can navigate the complexities of LLMs and RAG (Retrieval-Augmented Generation) while mentoring technical teams and communicating model trade-offs to stakeholders.

Responsibilities

- AI System Architecture : Lead the design and implementation of end-to-end AI solutions, integrating Machine Learning, Deep Learning, and LLMs to solve complex business challenges.

- Generative AI & RAG Orchestration : Architect advanced Retrieval-Augmented Generation (RAG) pipelines using vector databases like Milvus, FAISS, or Pinecone to enhance model context and accuracy.

- Advanced Model Development : Design and optimize deep learning models using PyTorch or TensorFlow, applying CNNs, RNNs, and Transformer-based architectures with attention mechanisms.

- Machine Learning Prototyping : Utilize NumPy, Pandas, and Scikit-learn for rapid ML prototyping, focusing on feature engineering, model selection, and ensemble methods like XGBoost and LightGBM.

- Model Fine-Tuning & Optimization : Lead the fine-tuning of Hugging Face Transformers, optimizing for tokenization, embeddings, and specific JSON schema outputs through function calling.

- Data Engineering Collaboration : Partner with data teams to build robust data wrangling pipelines and maintain high-performance SQL fundamentals for data extraction.

- DevOps & Containerization : Ensure seamless model deployment and reproducibility by leveraging Git/GitHub, Jupyter, and basic Docker containerization.

- Technical Thought Leadership : Stay abreast of cutting-edge AI research, experimenting with new optimization techniques (Adam, SGD) and regularization methods to maintain competitive advantage.

- Clear Communication : Act as a strategic liaison, explaining complex ML/LLM trade-offs to non-technical stakeholders to ensure alignment with business goals.

- Ownership & Mentorship : Take a proactive ownership mindset for project delivery while fostering a culture of curiosity and technical excellence within the team.

Technical Requirements :

- Core Programming : Advanced proficiency in Python, data structures, and algorithmic design.

- Deep Learning Mastery : Hands-on experience with Transformers, LSTMs, and optimization techniques like batch normalization and dropout.

- Generative AI Stack : Proven expertise in Prompt Engineering, tool calling, and managing high-dimensional vector embeddings.

- ML Foundations : Deep understanding of supervised/unsupervised learning, cross-validation, and regularization.

- Tools & Infrastructure : Working knowledge of Vector Databases, SQL, and basic MLOps using Docker.

Preferred Skills :

- LLM Specialization : Experience with open-source LLM fine-tuning and deployment.

- ElasticSearch : Knowledge of integrating search-based retrieval with AI models.

- Production Deployment : Proven track record of moving AI prototypes from Jupyter notebooks to scalable production environments.

Core Competencies :

- Applied AI Exposure : A solid academic foundation paired with strong, real-world AI/ML implementation experience.

- Experimental Curiosity : A relentless drive to learn and apply the latest advancements in Artificial Intelligence.

- Problem-Solving Mindset : Exceptional ability to deconstruct ambiguous business requirements into structured AI solutions.

- Result Driven : A focus on delivering tangible business impact through high-performance AI systems